Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
159
INVOLUNTARY OUTPATIENT
COMMITMENT: THE LIMITS OF
PREVENTION
Candice T. Player*
Preventive outpatient commitment laws require people with mental illnesses
to participate in mental health treatment before they meet the criteria for
inpatient civil commitment—clear and convincing evidence of mental illness and
dangerousness to self or others. These laws apply to people who are chronically
ill but not imminently dangerous. Most outpatient commitment laws do not
require a judicial determination of incompetence, nor do they require a criminal
charge or a criminal conviction. As such, outpatient commitment statutes unearth
an old question on law, ethics, and the limits of prevention: under what
circumstances can we impose substantial restraints on individual liberty because
we believe a person is likely to harm himself or others before he actually has
done so?
Although most authors rest the moral justification for outpatient commitment
on a mental impairment—be it impaired insight, decisional-incapacity or
incompetence to refuse treatment, this Article claims that government
interventions into self-regarding harm and other-regarding harm require distinct
moral justifications. When our primary concern is one of self-regarding harm, a
court order to participate in outpatient treatment may be appropriate, but only
for people with mental illnesses who are incompetent to make treatment decisions
on their own. If, however, we are concerned about harm to others, a court order
to participate in outpatient treatment may be appropriate, but only for people
with mental illnesses who lack the moral capacities for criminal responsibility—
either because they are unlikely to appreciate the wrongfulness of their conduct
or because they are unable to conform their conduct to the requirements of the
law.
INTRODUCTION....................................................................................................... 160 I. THE RISE OF OUTPATIENT COMMITMENT ......................................................... 166
A. Deinstitutionalization ................................................................................. 166 1. The (Broken) Promise of Community Mental Health ......................... 166
* Assistant Professor of Law, Northwestern University School of Law; Ph.D. Ethics and Health Policy, Harvard University 2013; J.D. Harvard Law School, 2009; M.Phil, Institute of Criminology, Cambridge University, 2003 A.B. Harvard College, 2002. I am grateful to Anita L. Allen, Anne Barnhill, Eric Beerbohm, Stephanos Bibas, Dan Brock, Glenn Cohen, Steve Joffe, Michelle Mello, Stephen Morse, Ted Ruger, and Tobias Wolff for their comments on this Article. All errors are my own.
160 STANFORD LAW & POLICY REVIEW [Vol. 26:159
2. A Changing Legal Landscape ............................................................. 169 3. “Rotting with Their Rights On” .......................................................... 172
B. Preventive Outpatient Commitment ........................................................... 175 1. New York ............................................................................................ 176
a. Mechanics ...................................................................................... 176 b. Program Implementation ............................................................... 178
2. California ............................................................................................ 179 3. North Carolina ..................................................................................... 181
C. Empirical Research on Effectiveness ......................................................... 182 1. Randomized Controlled Trials ............................................................ 182 2. Observational and Quasi-Experimental Designs ................................. 185
D. Constitutional Challenges .......................................................................... 187 II. MORAL JUSTIFICATIONS AND PUBLIC POLICY RATIONALES ............................. 191
A. Harm to Others........................................................................................... 191 1. Violence and Mental Illness ................................................................ 192
a. Community Surveys ...................................................................... 192 b. Psychosis and Violence ................................................................. 194
2. Risk Assessment and Violence ........................................................... 196 3. The Criminal-Civil Distinction ........................................................... 199
B. Harm to Self ............................................................................................... 201 1. Autonomy Defined .............................................................................. 204 2. Conceptual Hurdles and Replies ......................................................... 205 3. Autonomy and Mental Illness ............................................................. 208 4. When—If Ever—Is Paternalism Justified? ......................................... 210
a. Retrospective Endorsement Theories ............................................ 210 b. The Soft Paternalist Strategy ......................................................... 211
C. Impaired Insight ......................................................................................... 212 1. Impaired Insight Defined .................................................................... 212 2. The Neuroscience of Insight ............................................................... 213 3. Impaired Insight in The Courtroom..................................................... 216 4. An Objection from Autonomy ............................................................ 217
III. THE LIMITS OF PREVENTION ............................................................................. 218 A. Harm to Self ............................................................................................... 220
1. Competence to Refuse Treatment ....................................................... 220 a. Understanding and Appreciation ................................................... 221 b. Reasoning and Communication ..................................................... 225
2. Competence without Insight ................................................................ 228 B. Harm to Others........................................................................................... 229
1. Cognitive Impairment ......................................................................... 231 2. Volitional Impairment ......................................................................... 232
IV. OBJECTIONS AND CONCLUSIONS ...................................................................... 235
INTRODUCTION
In the years since deinstitutionalization, one of the most important
questions in mental health policy is this: how can we care for psychiatric
patients in the community who need treatment but resist treatment nonetheless?
2015] INVOLUNTARY OUTPATIENT COMMITMENT 161
The problem is particularly apparent for people with chronic mental illnesses
who are high utilizers of inpatient care—so called “revolving door patients.”
These patients improve when they are hospitalized and treated with
psychotropic medications but frequently stop taking their medications shortly
after they are released, creating a cycle of relapse and rehospitalization.1
Eventually, revolving door patients will deteriorate and meet the criteria for
inpatient civil commitment—clear and convincing evidence of mental illness
and dangerousness to self or others. However, involuntary outpatient
commitment laws permit courts to intervene and order people with mental
illnesses to comply with treatment in the community. A subtype of involuntary
outpatient commitment—known as preventive outpatient commitment—
permits court-ordered treatment for people with mental illnesses who do not
satisfy the criteria for inpatient commitment. Proponents of these laws tout
them as a solution to the revolving door problem.2
Forty-two states and the District of Columbia have involuntary outpatient
commitment laws.3 At least nine jurisdictions permit preventive outpatient
1. Joan B. Gerbasi et al., Resource Document on Mandatory Outpatient Treatment, 28 J. AM. ACAD. PSYCHIATRY & L. 127, 128 (2000). Many revolving door patients have been diagnosed with schizophrenia or schizoaffective disorder. These patients impose an enormous economic burden on the healthcare system due to the high cost of hospitalization. See Thomas W. Haywood et al., Predicting the “Revolving Door” Phenomenon Among Patients with Schizophrenic, Schizoaffective, and Affective Disorders, 152 AM. J. PSYCHIATRY 856 (1995); Patricia Thieda et al., An Economic Review of Compliance with Medication Therapy in the Treatment of Schizophrenia, 54 PSYCHIATRIC SERVICES 508 (2003); Peter J. Welden & Mark Olfson, Cost of Relapse in Schizophrenia, 21 SCHIZOPHRENIA BULLETIN 419 (1995).
2. See, e.g., Gerbasi et al., supra note 1; Ken Kress, An Argument for Assisted Outpatient Treatment for Persons with Serious Mental Illness Illustrated with Reference to a Proposed Statute for Iowa, 85 IOWA L. REV. 1269 (2000).
3. ALA. CODE § 22-52-10.2 (Westlaw through 2014 Sess.); ALASKA STAT. ANN. §
47.30.795 (West, Westlaw through 2014 Sess.); ARIZ. REV. STAT. ANN. § 36-540 (Westlaw through 2014 Sess.); ARK. CODE. ANN. § 20-47-214(b)(1) (West, Westlaw through 2014 Sess.); CAL. WELF. & INST. CODE § 5346 (West, Westlaw through 2014 Sess.) (“Laura’s Law”); COLO. REV. STAT. ANN § 27-65-111 (West, Westlaw through 2014 Sess.); DEL. CODE. ANN. tit. 16, § 5010(2) (West, Westlaw through 2014 Sess.); D.C. CODE § 21-545(b)(2)
(Westlaw through 2014 Sess.); FLA. STAT. ANN. § 394.4655 (West, Westlaw through 2014 Sess.); GA. CODE ANN. § 37-3-81.1(2) (West, Westlaw through 2014 Sess.); HAW. REV. STAT. § 334-121 (West, Westlaw through 2014 Sess.); IDAHO CODE ANN. § 66-339A (1998) (re-pealed 1998); 405 ILL. COMP. STAT. 5 / 3-751 (West, Westlaw through 2014 Sess.); IND. CODE ANN. § 12-26-14-1 (West, Westlaw through 2014 Sess.); IOWA CODE ANN. § 229.13
(West, Westlaw through 2014 Sess.); KAN. STAT. ANN. § 59-2967 (West, Westlaw through 2014 Sess.); KY. REV. STAT. ANN. § 202A.081(1) (West, Westlaw through 2014 Sess.); LA. REV. STAT. ANN. § 28:55(E)(1) (Westlaw through 2014 Sess.); ME. REV. STAT. ANN. tit. 34-B, § 3873-A (Westlaw through 2014 Sess.); MICH. COMP. LAWS ANN. § 330.1468 (West, Westlaw through 2014 Sess.) (“Kevin’s Law”); MINN. STAT. ANN. § 253B.065(5)(b) (West, Westlaw through 2014 Sess.); MISS. CODE ANN. § 41-21-73(4) (West, Westlaw through 2014 Sess.); MO. ANN. STAT. § 632.330(2) (West, Westlaw through 2014 Sess.); MONT. CODE
ANN. § 53-21-127(2) (West, Westlaw through 2013 Sess.1996); NEB. REV. STAT. ANN. § 71-909 (West, Westlaw through 2014 Sess.); NEV. REV. STAT. ANN. § 433A.310 (West, Westlaw through 2014 Sess.); N.H. REV. STAT. ANN. § 135-C:1 (Westlaw through 2014 Sess.); N.J.
162 STANFORD LAW & POLICY REVIEW [Vol. 26:159
commitment.4 Several states, including California, New Jersey, Florida, and
Michigan, have enacted involuntary outpatient commitment statutes of both
varieties in recent years. Although some states rarely use their outpatient
commitment laws, other states have been more aggressive.5 Since New York
established its outpatient commitment program in 1999, 12,218 New Yorkers
have been under a court order to participate in mental health treatment.6 The
shootings in Newtown, Connecticut; Aurora, Colorado; and Isla Vista,
California have prompted other states, including Connecticut, Nevada, and
Pennsylvania, to consider adopting outpatient commitment statutes or
strengthening existing provisions.7 In Congress, The Helping Families in
STAT. ANN. § 30:4-27.8a (West, Westlaw through 2014 Sess.) (“Gregory’s Law”); N.Y. MENTAL HYG. LAW § 9.60(C) (McKinney, Westlaw through 2014 Sess.) (“Kendra’s Law”); N.C. GEN. STAT. ANN. § 122C-263(d)(1) (West, Westlaw through 2014 Sess.); N.D. CENT. CODE § 25-03.1-20(1) (West, Westlaw through 2013); OHIO REV. CODE ANN. § 5122.15(C)
(West, Westlaw through 2014 Sess.); OKLA. STAT. ANN. tit. 43A, § 5-415E (West, Westlaw through 2014 Sess.); OR. REV. STAT. ANN. § 426.130(1)(a)(C)(ii) (West, Westlaw through 2014 Sess.); 50 PA. STAT. ANN § 7304 (West, Westlaw through 2014 Sess.); R.I. GEN. LAWS
ANN. § 40.1-5-8(j) (West, Westlaw through 2014 Sess.); S.C. CODE ANN. § 44-17-580
(Westlaw through 2013 Sess.); S.D. CODIFIED LAWS § 27A-9-10.1 (2000) (repealed 1991); TEX. HEALTH & SAFETY CODE ANN. § 574.034(b) (West, Westlaw through 2013 Sess.); UTAH
CODE ANN. § 62A-15-631 (West, Westlaw through 2014 Sess.); VT. STAT. ANN. tit. 18, §
7618 (West, Westlaw through 2014 Sess.); VA. CODE ANN. § 37.2-817 (West, Westlaw through 2014 Sess.); WASH. REV. CODE ANN. § 71.05.240 (West, Westlaw through 2014 Sess.); W. VA. CODE ANN. § 27-1-9 (West, Westlaw through 2014 Sess.); WIS. STAT. ANN. §
51.20 (West, Westlaw through 2013 Sess.); WYO. STAT. ANN. § 25-10-110(j) (West, Westlaw through 2014 Sess.). The six states without outpatient commitment laws are Connecticut, Maryland, Massachusetts, New Hampshire, New Mexico and Tennessee. See also Mark R. Munetz et al., Commentary: Capacity-Based Involuntary Outpatient Treatment, 28 J. AM. ACAD. PSYCHIATRY & L. 145 (2000) (discussing use of medication guardianship orders in Massachusetts, a state without an outpatient commitment statute).
4. CAL. WELF. & INST. CODE § 5346 (West, Westlaw through 2014 Sess.) (“Laura’s Law”); FLA. STAT. ANN. § 394.4655 (West, Westlaw through 2014 Sess.); GA. CODE ANN. §
37-3-81.1(2) (West, Westlaw through 2014 Sess.); HAW. REV. STAT. § 334-121 (West, Westlaw through 2014 Sess.); IDAHO CODE ANN. 66-339A (1998) (repealed 1998); MINN. STAT. ANN. § 253B.065(5)(b) (West, Westlaw through 2014 Sess.); N.Y. MENTAL HYG. LAW
§ 9.60(C) (McKinney, Westlaw through 2014 Sess.) (“Kendra’s Law”); N.C. GEN. STAT. ANN. § 122C-263(d)(1) (West, Westlaw through 2014 Sess.); TEX. HEALTH & SAFETY CODE
ANN. § 574.034(b) (West, Westlaw through 2013 Sess.).
5. See, e.g., John Petrila & Annette Christy, Florida’s Outpatient Commitment Law: A Lesson in Failed Reform?, 59 PSYCHIATRIC SERVICES 21, 21 (2008) (“In a state with a population that is approaching 19 million people, there have been a total of 71 orders for outpatient commitment in nearly three years.”); see also RISDON N. SLATE, JACQUELINE K. BUFFINGTON-VOLLUM & W. WESLEY JOHNSON, THE CRIMINALIZATION OF MENTAL ILLNESS: CRISIS AND OPPORTUNITY FOR THE JUSTICE SYSTEM 226 (2d ed. 2013).
6. N.Y. STATE OFFICE OF MENTAL HEALTH, ASSISTED OUTPATIENT TREATMENT
REPORTS [hereinafter N.Y. OMH AOT REPORTS], http://bi.omh.ny.gov/aot/ statistics?p=under-court-order (last visited Nov. 28, 2014).
7. See, e.g., NY Secure Ammunition and Firearms Enforcement (SAFE) Act, ch. 1, 2013 N.Y. Sess. Laws 1 (McKinney) [hereinafter SAFE Act] (expanding outpatient commitment law in New York); see also S. 452, Gen. Assemb., Feb. Sess. (Conn. 2012) (proposing involuntary outpatient commitment in Connecticut); H. 550, 197th Gen. Assemb., Reg. Sess. (Pa. 2013) (proposing an assisted outpatient treatment amendment to
2015] INVOLUNTARY OUTPATIENT COMMITMENT 163
Mental Health Crisis Act (H.R. 3717) would authorize $60 million dollars in
federal grants over four years to implement outpatient commitment programs.8
Although outpatient commitment orders were first introduced in the United
States, they are also an international phenomenon. Outpatient commitment laws
exist in Israel, Canada, Australia, and New Zealand.9 Amendments to the 1983
Mental Health Act introduced outpatient commitment orders to England and
Wales in November 2008.10
Several states have modeled their outpatient commitment statutes on New
York’s preventive outpatient commitment law, known as Kendra’s Law.11 New
York passed Kendra’s Law in memory of Kendra Webdale, a young woman
who was pushed to her death in front of an oncoming train by Andrew
Goldstein, a man with untreated schizophrenia. Under Kendra’s Law, a court
can order a person with a mental illness to participate in an “assisted outpatient
treatment” (AOT) program. A typical AOT order includes a host of
interventions designed to improve medication compliance in the community,
among them—periodic blood tests or urinalysis to determine compliance with
prescribed medications; counseling and toxicology screens for patients with a
history of substance abuse; group therapy, day or partial day programming; and
supervised living arrangements.12 For those who are not under a supervised
housing requirement, courts will sometimes order an assertive community
treatment (ACT) team to visit the patient’s home.
Much of the controversy surrounding outpatient commitment has focused
on the legal rights of people with mental disabilities and whether these laws are
an effective solution to hospital recidivism.13 Yet even if these laws are
effective and fall within the wide boundaries set by constitutional norms, hard
the Mental Health Procedures Act in Pennsylvania). See generally Arielle Levin Becker, After Newtown: A Rethinking of Connecticut Law on Outpatient Commitment, CONN. MIRROR (Jan. 2, 2013), http://ctmirror.org/after-newtown-rethinking-connecticut-law-outpatient-commitment [hereinafter Becker, After Newtown]; Arielle Levin Becker, A Judge’s Frustrations Lead to Proposal, Outcry from Advocates, CONN. MIRROR (Mar. 30, 2012), http://ctmirror.org/judges-frustrations-lead-proposal-outcry-advocates [hereinafter Becker, Judge’s Frustrations].
8. H.R. 3717, 113th Cong. (2013).
9. RACHEL CHURCHILL ET AL., KINGS COLL. LONDON, INST. OF PSYCHIATRY, INTERNATIONAL EXPERIENCES OF USING COMMUNITY TREATMENT ORDERS 28 (2007); see also, John Dawson, Fault-lines in Community Treatment Order Legislation, 29 INT’L J.L. &
PSYCHIATRY 482 (2006) (discussing outpatient commitment legislation in six British Commonwealth Jurisdictions).
10. The introduction of community treatment orders in England and Wales has sparked a heated debate. See generally Mark Taylor, Community Treatment Orders and Their Use in the UK, 16 ADVANCES IN PSYCHIATRIC TREATMENT 260 (2010).
11. N.Y. MENTAL HYG. LAW § 9.60 (a)(1) (McKinney, Westlaw through 2014 Sess.).
12. Id.
13. See, e.g., Michael Allen & Vicki Fox Smith, Opening Pandora’s Box: The Practical and Legal Dangers of Involuntary Outpatient Commitment, 52 PSYCHIATRIC
SERVICES 342 (2001); Jennifer Honig & Susan Stefan, New Research Continues to Challenge the Need for Outpatient Commitment, 31 NEW. ENG. J. ON CRIM & CIV. CONFINEMENT 109 (2005).
164 STANFORD LAW & POLICY REVIEW [Vol. 26:159
moral questions remain. The central concern of this Article can be thought of in
terms of the following question: under what circumstances can we impose
substantial restraints on individual liberty because we believe a person is likely
to harm himself or others before he has actually done so?
Supporters of Kendra’s Law rest the moral justification for intervention on
harm to self and others. Yet by itself, harm fails to provide a principled
distinction between people with mental illnesses and others who might also
refuse treatment. Consider the alcoholic who persists in driving drunk. We
could easily imagine a Kendra’s Law for people with substance abuse
disorders—replete with weekly Alcoholics Anonymous meetings, toxicology
tests, and home visits where ACT teams conduct “bottle checks” instead of pill
checks. Such a regime would pay dividends in preventable deaths due to
vehicular homicide, yet we do not have one. Instead we rely on the criminal
law to deter these harms.
Resting the argument for assisted outpatient treatment on harm to self
seems equally problematic. We do not require people with diabetes to take
medications that have the power to prevent blindness, amputation, coma, and
death. Nor do we require smokers to stop smoking any more than we require
people with cardiovascular disease or high cholesterol to participate in classes
on the dangers of inactivity and a fatty diet. As a general matter, courts do not
intervene in self-regarding treatment decisions, without a finding of
incompetence, no matter how grave the potential harm.
Supporters of outpatient commitment contend that quite unlike people with
diabetes, cardiovascular disease and high cholesterol, many people with major
mental illnesses like schizophrenia, bipolar disorder, and depression lack
insight into their illnesses, and when combined with a substantial risk of harm
to self or others, this lack of insight provides sufficient justification for court
ordered treatment, even when they do not meet the legal definition of
incompetence.14 In psychiatry the term “insight” refers to a person’s awareness
that he or she is suffering from a mental disorder.15 Conventional definitions of
competence to refuse treatment include insight as an element of competence. In
recent years, the notion that impaired insight provides a moral justification for
outpatient commitment has grown tremendously in popularity and influence. I
will argue that impaired insight fails to provide a strong justification for
outpatient commitment.
Most authors on bioethics and mental health law rest the moral justification
for outpatient commitment on a mental impairment—be it impaired insight,
decisional-incapacity, or incompetence to refuse treatment.16 What should we
14. See, e.g., Xavier F. Amador & Andrew A. Shiva, Insight into Schizophrenia: Anosognosia, Competency, and Civil Liberties, 11 GEO. MASON U. C.R. L.J. 25 (2000); E. Fuller Torrey & Mary Zdanowicz, Outpatient Commitment: What, Why, and for Whom?, 52
PSYCHIATRIC SERVICES 337 (2001).
15. See infra notes 329-34 and accompanying text.
16. See, e.g., Eric Dale, Is Supervised Community Treatment Ethically Justifiable?, 36 J. MED. ETHICS 271 (2009) (concluding that supervised community treatment is justifiable
2015] INVOLUNTARY OUTPATIENT COMMITMENT 165
say about a person who presents a substantial risk of harm to others by virtue of
mental illness but who is competent to refuse treatment nonetheless? One
response is to do nothing and claim that our hands are tied until the person
commits or attempts a crime.17 Below I argue that a court order to participate in
outpatient treatment may be permissible notwithstanding a finding of
incompetence. Our challenge, however, will be to distinguish persons with
mental disorders who are appropriately subject to preventive intervention
through outpatient commitment from others whose dangerous behaviors are
more appropriately controlled by the criminal law.
Part II.A sets the stage for discussion by providing a brief history of
institutionalization, deinstitutionalization, and their critics, with a focus on the
circumstances leading to Kendra’s Law in New York. Part II.B describes
preventive outpatient commitment laws in New York, California, and North
Carolina. Part II.C discusses empirical research on outpatient commitment,
while Part II.D examines constitutional challenges to outpatient commitment
laws. Part III opens the analytic section of this Article by refuting common
justifications for outpatient commitment. There is a place for outpatient
commitment, but in their current form outpatient commitment statutes violate
basic moral norms. To that end, Part IV develops a rights-based moral
framework for involuntary outpatient commitment.18 I begin by contrasting
consequentialist with non-consequentialist theories of moral justification. From
a utilitarian outlook, individual rights have no moral force aside from their
contribution to utility; however, the framework I present in Part IV is a
distinctly liberal one. It denies that moral rights can be infringed simply
because doing so would produce gains in utility.
for people with severe mental disorders who are incompetent to make treatment decisions); John Dawson & George Szmukler, Fusion of Mental Health and Incapacity Legislation, 188 BR. J. OF PSYCHIATRY 504 (2006) (proposing a single incapacity standard to govern involuntary treatment for physical and mental illnesses); Howard H. Goldman, Outpatient Commitment Reexamined: A Third Way, 65 PSYCHIATRIC SERVICES 816, 816 (2014) (proposing a “decisional capacity” standard for outpatient commitment); Michael A. Hoge & Elizabeth Grottole, The Case Against Outpatient Commitment. 28 J. AM. ACAD. PSYCHIATRY
& L. 165 (2000) (challenging the legitimacy of outpatient commitment for patients who are neither incompetent nor imminently dangerous); Mark R. Munetz, Commentary: Capacity-Based Involuntary Outpatient Treatment, 28 J. AM. ACAD. PSYCHIATRY & L. 145 (2000); Mark R. Munetz et al., The Ethics of Mandatory Community Treatment, 31 J. AM. ACAD. PSYCHIATRY & L. 173 (2003) (claiming that outpatient commitment may be justified for people who lack decision-making capacity); Elyn R. Saks, Involuntary Outpatient Commitment, 9 PSYCHOL. PUB. POL’Y & L. 94, 95 (2003) (proposing “one free shot” of preventive outpatient commitment when patients are incompetent to refuse treatment, but otherwise requiring a judicial determination of incompetence thereafter); Torrey & Zdanowicz, supra note 14, at 337 (describing impaired illness awareness as a moral justification for outpatient commitment).
17. See, e.g., Dawson & Smuzkler, supra note 16, at 504.
18. See RONALD DWORKIN, TAKING RIGHTS SERIOUSLY 171-72 (1977) (distinguishing rights-based moral theories from goal-based, or utilitarian, moral theories). I borrow the term “rights-based theory” from Dworkin.
166 STANFORD LAW & POLICY REVIEW [Vol. 26:159
Although most bioethicists claim that outpatient commitment orders are
justifiable if they are limited to people who are incompetent to refuse treatment
or otherwise cognitively impaired, one of the important claims in Part IV is that
government interventions into self-regarding harm and other-regarding harm
require distinct moral justifications. When our primary concern is one of self-
regarding harm, Part IV contends that a court order to participate in outpatient
treatment may be appropriate, but only for people with mental illnesses who are
unable to make competent treatment decisions on their own. In contrast to other
authors on competence, I argue that “appreciation,” the legal correlate of
insight, should have no role to play in our thinking about competence.
At times, we will also worry that a decision to refuse outpatient treatment
could not only result in harm to oneself, but also harm to others. Criminal law
scholars have long understood civil commitment as an alternative system of
social control for dangerous yet morally non-responsible, persons.19 Still, they
disagree on the indicia of moral non-responsibility. In Part IV, I argue that
when our primary concern is one of other-regarding harm, a court order to
participate in outpatient treatment may be appropriate, but only for people with
mental disorders who lack the moral capacities for criminal responsibility—
either because they are unlikely to appreciate the wrongfulness of their conduct
or because they are unable to conform their conduct to the requirements of the
law.
I. THE RISE OF OUTPATIENT COMMITMENT
A. Deinstitutionalization
1. The (Broken) Promise of Community Mental Health
During the first half of the twentieth century, civil commitment decisions
were predicated on the “best interests” of the patient and left in the hands of
physicians or family members.20 Most civil commitments were accomplished
with a two physician certificate, whereby patients were hospitalized on the
statement of two physicians that they were suffering from a mental disorder and
19. See, e.g., ROBERT F. SCHOPP, COMPETENCE, CONDEMNATION, AND COMMITMENT: AN INTEGRATED THEORY OF MENTAL HEALTH LAW (2001); Gary Watson, The Insanity Defense, in THE ROUTLEDGE COMPANION TO PHILOSOPHY OF LAW (Andrei Marmor, ed. 2012); Bruce Winick, Civil COMMITMENT: A THERAPEUTIC JURISPRUDENCE MODEL 60 (2005); Adam J. Falk, Sex Offenders, Mental Illness and Criminal Responsibility: The Constitutional Boundaries of Civil Commitment after Kansas v. Hendricks, 25 AM. J. L. &
MED. 117 (1999); Stephen J. Morse, Rationality and Responsibility, 74 S. CAL. L. REV. 251 (2000); Paul H. Robinson, Foreword: The Criminal-Civil Distinction and Dangerous Blameless Offenders, 83 J. CRIM. L. & CRIMINOLOGY 693 (1993); Stephen J. Schulhofer, Two Systems of Social Protection: Comment on the Civil-Criminal Distinction, with Particular Reference to Sexually Violent Predator Laws, 7 J. CONTEMP. LEGAL ISSUES 69 (1996).
20. PAUL S. APPELBAUM, ALMOST A REVOLUTION: MENTAL HEALTH LAW AND THE
LIMITS OF CHANGE 18-22, 129 (1994).
2015] INVOLUNTARY OUTPATIENT COMMITMENT 167
in need of care or treatment.21 In most states, commitment could be achieved
without a hearing, without counsel, and without a legal remedy, aside from a
writ of habeas corpus.22 By the mid-1950s, when the number of
institutionalized psychiatric patients reached its peak, more than 550,000
inpatients resided in state mental hospitals.23 By the mid-1980s, however,
fewer than 120,000 psychiatric patients resided in state hospitals.24
Several factors encouraged a shift toward community mental health care.
During the late 1950s, attitudes toward institutional psychiatry began to
change.25 Social scientists questioned the ability of psychiatrists to diagnose
mental illness reliably26 while labeling theorists27 and radical anti-
psychiatrists28 insisted that psychiatric diagnoses were no more than
convenient labels designed to suppress nonconforming behavior. So labeled,
persons deemed mentally ill would in turn reproduce more disturbed behavior.
A further critique of psychiatry came from the civil rights movement. Civil
rights organizations argued that inpatient commitment standards were vague,
overbroad, and void for failure to consider less restrictive alternatives to
involuntary hospitalization.29 A third critique concerned the benefits of long-
term hospitalization.30 In the years following World War II, a series of exposés
called attention to deplorable conditions in state hospitals.31 For the first time,
21. See N.Y. MENTAL HYG. LAW § 9.27(a) (McKinney, Westlaw through 2014 Sess.) (authorizing non-emergency involuntary civil commitment if two physicians examine the committee and certify that he or she is mentally ill and “in need of involuntary care and treatment.”); see also AM. BAR FOUND., THE MENTALLY DISABLED AND THE LAW 63-65 (1961).
22. ALBERT DEUTSCH, THE MENTALLY ILL IN AMERICA: A HISTORY OF THEIR CARE
AND TREATMENT FROM COLONIAL TIMES 414, 436 (1949).
23. H. Richard Lamb, Deinstitutionalization at the Beginning of the New Millenium, in DEINSTITUTIONALIZATION: PROBLEMS AND PROMISE 3, 3 (Richard H. Lamb & Linda E. Weinberger eds., 2001).
24. GERALD N. GROB, THE MAD AMONG US: THE HISTORY OF THE CARE OF AMERICA’S
MENTALLY ILL 291 (1994).
25. APPELBAUM, supra note 20, at 4-10. 26. Symposium, On Being Sane in Insane Places, 13 SANTA CLARA LAW. 379 (1973).
27. R.D. LAING, THE POLITICS OF EXPERIENCE 78-79 (1967) (“Those who are diagnosed as schizophrenic are not ill but are reacting in a sane and rationale way to the intolerable emotional pressures placed on them by society and their families.”); THOMAS SCHEFF, BEING
MENTALLY ILL: A SOCIOLOGICAL THEORY (Aldine 1966).
28. See THOMAS SZAZ, THE MYTH OF MENTAL ILLNESS: FOUNDATIONS OF A THEORY OF
PERSONAL CONDUCT (1961).
29. Lessard v. Schmidt, 349 F. Supp. 1078, 1093 (E.D. Wis. 1972).
30. See generally ERVING GOFFMAN, ASYLUMS: ESSAYS ON THE SOCIAL SITUATION OF
MENTAL PATIENTS AND OTHER INMATES (1961) (describing psychiatric hospitals as “total institutions” in which people with mental disorders lose their identities and take on the de-humanizing, and ultimately debilitating, role of “patient”)
31. See ALBERT DEUTSCH, THE SHAME OF THE STATES (1948).
168 STANFORD LAW & POLICY REVIEW [Vol. 26:159
the emergence of psychotropic medications also offered the possibility of
treating people with mental illnesses in the community.32
Changes in federal policy accelerated the transition to community mental
health care. In 1963, President Kennedy called for “a bold new approach” to
treating mental illness, one that relied upon “new knowledge and new drugs,”
to treat people in communities rather than hospitals.33 The centerpiece of his
approach was the Community Mental Health Act.34 The Act provided $150
million in federal grants for constructing community mental health centers
(CMHCs).35 Two years later, a second federal law provided grants for staffing
mental health centers.36 Equally important were amendments to the Social
Security Act, resulting in the creation of Medicare and Medicaid.37 When
Congress passed Medicaid in 1965, the federal government excluded Medicaid
payments for psychiatric services rendered in state hospitals. In response, states
discharged large numbers of former inpatients to nursing homes and other
congregate care arrangements where Medicaid reimbursement was available.38
A few years later, Supplemental Security Income (SSI) provided financial
support for people with mental illnesses.39
While some former inpatients did well in the community, many others did
not. Far fewer community mental health centers were created than anticipated,
making it difficult for former inpatients to access care.40 The Community
Mental Health Act aimed to create 1500 mental health centers nationwide, but
by 1980, only 754 centers received federal funding.41 Community mental
health centers (CMHCs) were intended to replace state mental hospitals, but the
regulations governing CMHCs made no mention of how these centers would
coordinate with state hospitals or provide aftercare for former patients.42 The
inpatient populations of state mental hospitals declined throughout the 1960s
and 1970s; however, the decline was largely unrelated to community mental
health centers. Most CMHCs served a very different population.43 CMHCs
32. EDWARD SHORTER, A HISTORY OF PSYCHIATRY: FROM THE ERA OF THE ASYLUM TO
THE AGE OF PROZAC 279-80 (1997).
33. President John F. Kennedy, Speech to the 88th Congress: Special Message to Congress on Mental Illness and Mental Retardation, available at http:// www.jfklibrary.org/Asset-Viewer/Archives/JFKPOF-052-012.aspx.
34. Pub. L. No. 88-164, 77 Stat. 282 (1963) (also known as the Mental Retardation Facilities and Community Mental Health Centers Construction Act of 1963).
35. GROB, supra note 24, at 258.
36. Mental Retardation Facilities and Community Mental Health Centers Construction Act Amendments of 1965, Pub. L. No. 89-105., 77 Stat. 282 (1965).
37. The Social Security Amendments of 1965, Pub. L. No. 89-97, 79 Stat. 286 (1965).
38. APPELBAUM, supra note 20, at 50-51.
39. Social Security Amendments of 1972, Pub. L. No. 92-603, 86 Stat. 1329 (1972).
40. GROB, supra note 24, at 283-87.
41. Id. at 262.
42. E. FULLER TORREY, AMERICAN PSYCHOSIS HOW THE FEDERAL GOVERNMENT
DESTROYED THE MENTAL ILLNESS TREATMENT SYSTEM 62 (2013).
43. GROB, supra note 24, at 264.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 169
generally catered to the interests of the affluent “worried well,” namely,
patients with emotional problems, and frequently personal problems, rather
than people with severe mental disorders who were being discharged from state
hospitals.44
2. A Changing Legal Landscape
At the same time, the criteria for involuntary civil commitment were also
changing. In Lessard v. Schmidt, the Federal District Court for the Eastern
District of Wisconsin issued a landmark opinion on civil commitment law.45
Prior to Lessard, Wisconsin law permitted a judge or jury to order civil
commitment if the court was satisfied that the person was mentally ill or infirm
and “a proper subject for custody and treatment” based on a preponderance of
the evidence.46 Lessard held that the statute was unconstitutional insofar as it
permitted civil commitment without proof of mental illness and dangerousness
to self or others beyond a reasonable doubt. The district court held that states
must prove dangerousness based on “a recent overt act, attempt or threat to do
substantial harm to oneself or another.”47 In addition, Lessard required states to
establish that there was an extreme likelihood that the person would do
“immediate harm to himself or others” if not confined.48 The district court also
mandated a panoply of procedural due process protections, including timely
and effective notice of the “charges” justifying detention, the right to a jury tri-
al, the right to adversary counsel, and the privilege against self-incrimination.
Several years later, the Supreme Court heard substantive and procedural
due process challenges to civil commitment in O’Connor v. Donaldson49 and
Addington v. Texas.50 Donaldson began when Kenneth Donaldson brought suit,
alleging that the Florida State Hospital and its superintendent intentionally and
maliciously deprived him of his liberty.51 Uncontroverted testimony adduced at
trial established that Donaldson was not dangerous and that the hospital was not
treating his illness but instead had only provided non-curative “milieu
therapy.”52 In a unanimous opinion, the Supreme Court held that “a finding of
‘mental illness’ alone cannot justify a State’s locking a person up against his
will and keeping him indefinitely in simple custodial confinement.”53
44. TORREY, supra note 42, at 78; see also, RAEL JEAN ISAAC & VIRGINIA C. ARMAT, MADNESS IN THE STREETS: HOW PSYCHIATRY AND THE LAW ABANDONED THE MENTALLY ILL 95-96 (1990).
45. Lessard v. Schmidt, 349 F. Supp. 1078, 1093 (E.D. Wis. 1972).
46. Id. at 1093 (citation omitted).
47. Id.
48. Id.
49. 422 U.S. 563 (1975).
50. 441 U.S. 418 (1979).
51. Donaldson, 422 U.S. at 565.
52. Id. at 569 (citation omitted).
53. Id. at 575.
170 STANFORD LAW & POLICY REVIEW [Vol. 26:159
“[W]ithout more,” Justice Stewart wrote, a state cannot confine “a
nondangerous individual who is capable of surviving safely in freedom” on his
own or with assistance from friends and family.54
The Supreme Court turned to the standard of proof for civil commitment
proceedings in Addington v. Texas.55 Addington claimed that civil commitment
by less than proof beyond a reasonable doubt violated his right to procedural
due process. In an opinion by Chief Justice Burger, the Supreme Court held
that while the individual interest in the outcome of a civil commitment
proceeding is of sufficient gravity to require more than a preponderance of the
evidence, the Fourteenth Amendment requires no more than “clear and
convincing evidence.”56 Given the uncertainty of prediction and the fallibility
of diagnosis, the Chief Justice concluded that proof beyond a reasonable doubt
would impose a standard that states cannot meet.57
In the years following Lessard, Donaldson, and Addington, states modified
their civil commitment laws. By the end of the 1970s, every state raised the
54. Id. at 576. The Court did little to clarify the meaning of the phrase “without more,” leading some observers to conclude that, by itself, Donaldson would not preclude the involuntary civil commitment of a nondangerous person with a mental illness if the state also provided treatment. Indeed, Justice Stewart wrote, “[t]here is accordingly, no occasion in this case to decide whether the provision of treatment, standing alone, can ever constitutionally justify involuntary confinement . . . . In its present posture this case involves not involuntary treatment but simply involuntary custodial confinement.” Donaldson, 422 U.S. at 574 n.10 (1975). That language, along with the phrase “without more,” has led some commentators to conclude that the Court did not announce a dangerousness requirement in Donaldson, and therefore, inpatient civil commitment does not require proof of dangerousness. See, e.g., Geoffrey Linburn, Donaldson Revisited: Is Dangerousness a Constitutional Requirement for Civil Commitment?, 26 J. AM. ACAD. PSYCHIATRY L. 343 (1998); BRIAN STETTIN, TREATMENT ADVOCACY CTR., MENTAL HEALTH COMMITMENT LAWS: A SURVEY OF THE
STATES (2014), http://www.tacreports.org/storage/documents/2014-state-survey-abridged.pdf; Jeffrey Geller & Jonathan Stanley, Settling Doubts About the Constitutionality of Outpatient Commitment, 31 NEW ENG. J. ON CRIM. & CIV. CONFINEMENT 127, 132 (2005). Nonetheless, in subsequent opinions, the Supreme Court has clearly stated that civil commitment requires clear and convincing evidence of both mental illness and dangerousness. See, e.g., Jones v. United States, 463 U.S. 354, 362 (1983) (interpreting Addington v. Texas, 441 U.S. 418, 426-27 (1979), as holding that “the Due Process Clause requires the Government in a civil-commitment proceeding to demonstrate by clear and convincing evidence that the individual is mentally ill and dangerous”); Foucha v. Louisiana, 504 U.S. 71, 80 (1992) (citation omitted) (“The State may . . . confine a mentally ill person if it shows by ‘clear and convincing evidence that the individual is mentally ill and dangerous.’”).
55. 441 U.S. 418 (1979).
56. Id. at 433.
57. Id. at 430. Addington left the term “clear and convincing evidence” undefined; however, lower courts have defined clear and convincing evidence as evidence that makes the existence of a fact “highly probable,” or “much more probable than its falsity.” Ausch v. St. Paul Fire & Marine Ins. Co., 511 N.Y.S.2d. 919, 921-22 (N.Y. App. Div. 1987) (approving jury instruction that used “highly probable” language to describe clear and convincing evidence); Chenango Mut. Ins. Co. v. Charles, 652 N.Y.S.2d 134, 136 (N.Y. App. Div. 1997) (specifying that clear and convincing evidence standard requires “high probability”).
2015] INVOLUNTARY OUTPATIENT COMMITMENT 171
threshold for inpatient civil commitment from mental illness and the need for
treatment based on a preponderance of the evidence to clear and convincing
evidence of mental illness and dangerousness to self or others.58 Although not
required by Donaldson or Addington, several states required proof of
dangerousness to self or others based on a “recent overt act”59 or a substantial
risk of “imminent harm”60 in accordance with the district court decision in
Lessard v. Schmidt. In the same way, states mandated procedural due process
protections, including notice, the right to counsel, the right to confront and
cross-examine witnesses, and the right to appeal.61
58. APPELBAUM, supra note 20, at 28 (1994). The precise formulation of dangerousness to self or others varied from state to state. Some states implemented a “grave disability” standard as an alternative to the more rigorous “dangerousness to self standard.” See, e.g., Lanterman-Petris-Short Act, CAL. WELF. & INST. CODE §§ 5008(h)(1), 5150, 5250 (West, Westlaw through 2013-2014 Sess.) (allowing civil commitment for a person who “as a result of a mental disorder, is a danger to others, or to himself, or gravely disabled,” where “gravely disabled” shall mean that the person is “unable to provide for his basic personal needs for food, clothing and shelter”). Other states regarded a “substantial risk” or “likelihood” of serious harm as sufficient to meet the dangerousness requirement. See, e.g., N.Y. MENTAL HYG. LAW § 9.37 (McKinney, Westlaw through 2014 Sess.) (permitting involuntary hospitalization of a person whose mental illness is “likely to result in serious harm to himself or others,” where a “likelihood of serious harm” shall mean: (1) “substantial risk of physical harm to himself as manifested by threats of or attempts at suicide or serious bodily harm or other conduct demonstrating that he is dangerous to himself,” or (2) “a substantial risk of physical harm to other persons as manifested by homicidal or other violent behavior by which others are placed in reasonable fear or serious physical harm”).
59. See 1975 ALA. ACTS 2562, 2566 (Act No. 1226, § 10); CAL. WELF. & INST. CODE §§ 5260, 5300 (West, Westlaw through 2013-2014 Sess.); MASS. GEN. LAWS ANN. ch. 123, § 1 (West, Westlaw through 2014 Sess.); NEB. REV. STAT. ANN. § 71-908 [ME: Law transferred to 71-908](West, Westlaw through 2014 Sess.); N.C. GEN. STAT. § 122C-268 (West, Westlaw through 2014 Sess.); WASH. REV. CODE ANN. § 71.05.020 (West, Westlaw through 2014 Sess.); 1975 WIS. STAT. ANN. § 51.20 (West, Westlaw through 2013 Sess.); see also Commonwealth ex rel. Finken v. Roop, 339 A.2d 764, 778-79 (1975) (specifying that commitment requires “the commission of some overt act”); Doremus v. Farrell, 407 F. Supp. 509, 514-15 (D. Neb. 1975) (holding that due process and equal protection require that “the threat of harm has been evidenced by a recent overt act or threat.”); JOHN PARRY, CIVIL
MENTAL DISABILITY LAW, EVIDENCE AND TESTIMONY: A COMPREHENSIVE REFERENCE
MANUAL FOR JUDGES AND MENTAL DISABILITY PROFESSIONALS 476-77 (2010) (discussing state trend to eliminate imminence and overt act requirements).
60. See, e.g., CAL. WELF. & INST. CODE §§ 5260(a), 5300 (West, Westlaw through 2013-2014 Sess.); GA. CODE ANN. § 37-3-1(9.1) (West, Westlaw through 2014 Sess.); HAW. REV. STAT. ANN. § 334-60.2 (West, Westlaw, through2014 Sess.); MONT. CODE ANN. § 53-21-126. (West, Westlaw through 2013 Sess.); OHIO REV. CODE ANN. § 5122.01 (West, Westlaw through 2013-2014 Sess.); VA. CODE. ANN. § 37.2-814 (West, Westlaw through 1976).
61. See, e.g., ARIZ. REV. STAT. ANN. §§ 36-536-540 (Westlaw through 2014 Sess.) (providing notice, counsel, hearing); CAL. WELF. & INST. CODE § 5350 (West, Westlaw through 2014 Sess.) (providing notice, counsel, hearing, and jury trial); TEX. REV. CIV. STAT. ANN. art. 5547, §§ 40-57 (Vernon 1958 & Supp. 1978)(providing, at the time, notice, counsel, jury trial, and right to appeal).
172 STANFORD LAW & POLICY REVIEW [Vol. 26:159
In a series of cases beginning with Rennie v. Klein62 and Rogers v. Okin,63
psychiatric inpatients also gained the right to refuse anti-psychotic treatment.
Prior to Rennie and Rogers, hospitals assumed that involuntary commitment
extinguished the patient’s liberty interests, including the right of a competent
person to refuse treatment.64 However, in both cases, courts concluded that
committed patients were presumptively competent to make treatment
decisions.65 Therefore, without a legal determination of incompetence, patients
could not be medicated, except in an emergency when failing to do so would be
likely to result in substantial harm to the patient or others.66
3. “Rotting with Their Rights On”
As states began to modify the laws governing inpatient hospitalization,
critics argued that the new rules went too far. Psychiatrists began to describe
cases in which people with serious mental illnesses were not hospitalized either
because they were not “imminently or potentially dangerous,” or they lacked
suicidal intent.67 In a recurrent theme, critics argued that people with serious
mental illnesses were in effect “rotting with their rights on.”68
By the late 1980s, the well-publicized cases of Joyce Brown and Larry
Hogue came to symbolize the failures of the public mental health system. Joyce
Brown, who preferred the name “Billie Boggs,” first appeared on the streets of
Manhattan in 1987.69 Boggs lived on the corner of Second Avenue and 65th
Street, urinating and defecating on the sidewalk, burning dollar bills, and
screaming obscenities when assistance was offered.70 In October 1987, Ms.
Boggs was picked up by a local program designed to remove people with
mental illnesses from the streets when their lives were threatened by severe
weather and hypothermia.71 When the program attempted to hospitalize her,
62. 653 F.2d 836 (3d Cir. 1981).
63. 634 F.2d 650 (1st Cir. 1980).
64. Rennie, 653 F.2d at 843; Rogers, 634 F.2d at 657.
65. Rennie, 653 F.2d at 846; Rogers, 634 F.2d at 658.
66. Rennie, 653 F.2d at 852; Rogers, 634 F.2d at 654.
67. See Darold A. Treffert, Dying with Their Rights on, 2 PRISM 47, 49 (1974); Paul Chadoff, The Case for Involuntary Hospitalization of the Mentally Ill, 133 AM. J. PSYCHIATRY 496 (1976).
68. See Paul Appelbaum, “Rotting with Their Rights On”: Constitutional Theory and Clinical Reality in Drug Refusal by Psychiatric Patients, 7 BULL. AM. ACAD. PSYCHIATRY &
L. 306 (1979); Paul S. Appelbaum & Thomas G. Gutheil, The Boston Hospital Case: “Involuntary Mind Control, The Constitutional Right to Rot,” 75 AM. J. PUB. HEALTH, 722 (1985). Psychiatrists were not uniformly resistant to changes in civil commitment law. Cf., Seymour L. Halleck, A Troubled View of Current Trends in Forensic Psychiatry, 2 J. PSYCHIATRY & L. 135, 147 (1974) (“Many of us have long anguished over the power we hold over patients and have welcomed greater community control of that power.”).
69. GROB, supra note 24, at 302.
70. Jeanie Kasindorf, The Real Story of Billie Boggs: Was Koch Right or the Civil Libertarians?, N.Y. MAG., May 2, 1988, at 36, 37.
71. Id. at 41.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 173
Ms. Boggs filed a lawsuit with assistance from the New York Civil Liberties
Union.72
The psychiatrists who testified during her trial disagreed as to whether Ms.
Boggs was mentally ill.73 Even if Boggs were mentally ill, the trial court held
that the hospital failed to establish a “substantial risk of physical harm.”74
There was no evidence that Boggs was suicidal, and she was able to meet her
essential needs for food, clothing, and shelter.75 “It cannot be reasoned that
because Joyce Brown is homeless[,] she is mentally ill,” the court wrote.76
“What must be proved is that because she is mentally ill she is incapable of
providing herself with food, clothing, and shelter.”77 Boggs used the money she
received from panhandling to buy one meal each day, and the doctors for the
petitioner agreed that she was not malnourished and in “good physical
condition.”78
The Appellate Division reversed, finding the trial court’s conclusions as to
mental illness to be against the weight of evidence.79 In a vigorous dissent, two
members of the court agreed that Ms. Boggs was mentally ill but found the
evidence on dangerousness to be “highly questionable.”80 Ms. Boggs had never
assaulted anyone; she stayed warm by lying next to an air vent; and she was
never observed to be physically ill.81 Ms. Boggs was released from the hospital
one month later when a court held that she could not be medicated against her
will.82 Thereafter, Boggs enjoyed a brief stint as a national celebrity. She
appeared on 60 Minutes and Donahue, and in February 1988, she appeared as a
guest speaker at Harvard Law School.83 Yet only a few weeks later, the
symptoms of her psychosis reappeared, and Boggs was once again panhandling
on the streets of Manhattan.84
Several years later, Larry Hogue gained notoriety among New Yorkers as
“the Wild Man of 96th Street.”85 Hogue epitomized the revolving door problem
72. Id. at 37.
73. In re Boggs, 522 N.Y.S.2d 407, 410 (N.Y. Sup. Ct. 1987).
74. Id. at 411.
75. Id. at 412.
76. Id.
77. Id.
78. Id.
79. Boggs v. N.Y.C. Health & Hosp. Corp., 523 N.Y.S.2d 71, 86 (N.Y. App. Div. 1987).
80. Id. at 91.
81. Id. at 92.
82. See Matter of Boggs, 70 N.Y.2d 981 (1988); Josh Barbanel, Joyce Brown Obtains a Ban On Medicine, N.Y. TIMES, Jan. 16, 1988, at 29.
83. Kasindorf, supra note 71, at 44.
84. Rick Hampson, After a Brief Encounter with Fame, Homeless Billie Boggs Fades Away, FREE LANCE STAR, June 3, 1991, at 17.
85. Leigh Remizowski et al., The Wild Man of 96th St., Larry Hogue, Caught by Police on Upper West Side, N.Y. DAILY NEWS, May 30, 2009, http:// articles.nydailynews.com/2009-05-30/local/17922790_1_96th-larry-hogue-drug-habit.
174 STANFORD LAW & POLICY REVIEW [Vol. 26:159
and, for opponents of civil commitment reform, his case also underscored the
need for an outpatient commitment law.86 For years, Hogue terrorized New
Yorkers on Manhattan’s Upper West Side by siphoning gasoline from parked
cars, igniting newspapers soaked with gas, and then stuffing them into
tailpipes.87 Over the years, witnesses also observed Hogue jumping into
oncoming traffic and threatening people on the street with a nail-studded
club.88 In 1988, Hogue was convicted of reckless endangerment for pushing a
teenage girl in front of an oncoming truck.89 Yet because these incidents never
resulted in serious injury, Hogue never spent more than a year in jail.90 As a
patient at Creedmoor Hospital, Hogue was diagnosed with bipolar disorder,
crack addiction, and a traumatic brain injury.91 According to his doctors, when
Hogue was hospitalized and no longer abusing drugs, he was calm and amiable.
Yet, when he was no longer a danger to others, hospitals could no longer hold
him against his will.92 Following his release, Hogue would inevitably stop
taking his medications and revert to using drugs, leading him to behave in ways
that once again rendered him a danger to himself and others and subject him to
rehospitalization.93
By the early 1990s, the idea for an outpatient commitment law was already
well underway in the New York State Legislature. In 1994, the Legislature
established a pilot outpatient commitment program at Bellevue Hospital in
lower Manhattan.94 In a twist of fate, Andrew Goldstein visited the psychiatric
emergency room of Bellevue Hospital on at least two occasions in 1998, during
the tenure of the pilot program, complaining of auditory hallucinations and
sleep deprivation.95 After a few days in the hospital, Goldstein was released.96
A subsequent investigation into the quality of care Andrew Goldstein received
found that Goldstein repeatedly sought help for hallucinations and delusions.97
In the two-year period between early 1997 and January 1999, Goldstein
voluntarily admitted himself to state hospitals no fewer than thirteen times.98
86. E. Fuller Torrey, “The System Is Crazier Than He Is,” CITY JOURNAL (1998).
87. Seltzer v. Hogue, 594 N.Y.S.2d 781, 782 (N.Y. App. Div. 1993).
88. Remizowski, supra note 85.
89. Richard Perez-Pena, Crack Addict Found Able to Face Trial, N.Y. TIMES, Dec. 8, 1992, at B3.
90. Id.
91. Hogue, 594 N.Y.S.2d at 783.
92. Richard Perez-Pena, Mentally Ill Man Who Abused Drugs Is Freed, N.Y. TIMES, Feb. 5, 1993, Sec. B2.
93. GROB, supra note 24, at 304.
94. N.Y. MENTAL HYG. LAW § 41.55 (McKinney, Westlaw through 2014 Sess.).
95. See Michael Winerip, Bedlam on the Streets, N.Y. TIMES, May 23, 1999, at 42.
96. Id.
97. N.Y. STATE COMM’N ON QUALITY OF CARE FOR THE MENTALLY DISABLED, IN THE
MATTER OF DAVID DIX: A REPORT BY THE NEW YORK STATE COMMISSION ON QUALITY OF
CARE FOR THE MENTALLY DISABLED AND THE MENTAL HYGIENE MEDICAL REVIEW BOARD (1999).
98. Id.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 175
On more than one occasion, Goldstein requested long-term hospitalization at
Creedmoor.99 However, more often than not, he was turned down. Under
tremendous pressure to cut costs and reduce the number of inpatients, the
hospital could do little more than place Goldstein on a waiting list.
Several months later, in January 1999, Andrew Goldstein pushed Kendra
Webdale in front of an oncoming subway train. In Albany, former Attorney
General Elliot Spitzer seized the opportunity to create a permanent outpatient
commitment program in New York. In a statement to the press, Spitzer alluded
to the problems associated with deinstitutionalization:
It is clear that the law must be changed to protect both the public and the
mentally ill from danger . . . . The movement to deinstitutionalize has proven
to be a double-edged sword. Most individuals can and do function well in
society, but others with severe mental illness who are not taking their
prescribed medication can be a serious threat to themselves and the public.100
Three months later Julio Perez, a homeless man suffering from paranoid
schizophrenia, pushed Edgar Rivera into the path of an oncoming train,
severing both of his legs.101 According to his attorney, Perez harbored a
delusional belief that a conspiratorial network of Mexican assassins was trying
to kill him. Perez pushed Rivera, believing Rivera to be part of that network.102
In recent years, the shootings in Newtown, Connecticut and Isla Vista,
California have led to a renewed interest in preventive outpatient treatment.103
B. Preventive Outpatient Commitment
Preventive outpatient commitment is only one of at least three types of
outpatient commitment. The first, conditional release, applies to patients who
have been hospitalized and released on the condition that they will comply with
a treatment plan in the community. Outpatient commitment might also be used
as a less restrictive alternative to hospitalization for patients who meet the
criteria for inpatient commitment but who can be treated safely in the
99. Id.
100. Eliot Spitzer, Is Kendra’s Law the Answer?; Yes; Adversary System, 221 N.Y. L.J. 1 (1999); Press Release, N.Y. State Office of the Attorney Gen., Spitzer Proposes Mental Illness Legislation (Jan. 28, 1999), http://www.ag.ny.gov/press-release/spitzer-proposes-mental-illness-legislation; see also Press Release, N.Y. State Office of the Attorney Gen., Speaker Silver Joins Attorney General Spitzer in Calling for Passage of Kendra’s Law, http://www.ag.ny.gov/press-release/speaker-silver-joins-attorney-general-spitzer-calling-passage-kendras-law.
101. Andrew Jacobs, Man Is Pushed in the Subway and Loses Legs, N.Y. TIMES, Apr. 29, 1999, at A1.
102. The Associated Press, Jury Finds Subway Pusher Guilty of Attempted Murder, N.Y. TIMES, Oct. 22, 2000, at 239.
103. See, e.g., Becker, After Newtown, supra note 7; Conn. Debates Mandatory Outpatient Treatment for Mentally Ill (NPR Broadcast, Jan. 29, 2013); Siobhan Hughes, Isla Vista Rampage Reanimates Debate over Mental-Health Bill, WALL ST. J., May 28, 2014, http://online.wsj.com/articles/isla-vista-rampage-reanimates-debate-over-mental-health-bill-1401321183.
176 STANFORD LAW & POLICY REVIEW [Vol. 26:159
community nonetheless. Preventive outpatient commitment laws are a third,
and far more controversial, form of commitment. Preventive outpatient
commitment laws require people with mental illnesses to participate in
treatment even though they do not meet the criteria for inpatient commitment.
Part II provides a brief description of three well-known preventive outpatient
commitment statutes in New York, California, and North Carolina.
1. New York
a. Mechanics
Under Kendra’s Law a court may order a person who is eighteen years of
age or older to comply with an assisted outpatient treatment plan if the court
finds by clear and convincing evidence that the subject of the treatment plan
meets the following criteria. He or she must be suffering from a mental illness
and “unlikely to survive safely in the community without supervision, based on
a clinical determination.”104 The court must also find that a history of treatment
noncompliance has either: (i) been a significant factor leading to hospitalization
at least twice within the last thirty-six months, or (ii) resulted in one or more
acts of violent behavior toward self or others within the last forty-eight months,
or at least a threat or attempt at serious physical harm to self or others within
the last forty-eight months.105 In addition, the petitioner must provide clear and
convincing evidence that the subject of the petition is unlikely to participate in
outpatient treatment voluntarily “as a result of his or her mental illness.”106
Assisted outpatient treatment must be necessary to prevent a relapse or
deterioration, “which would be likely to result in serious harm to the person or
others.”107 Finally, the person must be likely to benefit from treatment, and
assisted outpatient treatment must be the least restrictive form of treatment
available.108 In some states, an assisted outpatient treatment order requires
evidence that the person lacks the capacity to make a treatment decision.109
However, Kendra’s Law does not include an incapacity requirement.110
104. N.Y. MENTAL HYG. LAW § 9.60(c)(2)-(3) (McKinney, Westlaw through 2014 Sess.).
105. Id. § 9.60(c)(4)(i)-(ii).
106. Id. § 9.60(c)(5).
107. Id. § 9.60(c)(6).
108. Id. § 9.60(c)(7), (h)(4).
109. See, e.g., HAW. REV. STAT. § 334-121(5) (West, Westlaw through 2014 Sess.) (requiring evidence that “[t]he person’s current mental status or the nature of the person’s disorder limits or negates the person’s ability to make an informed decision to voluntarily seek or comply with recommended treatment”); N.C. GEN. STAT. ANN. § 122C-271(a) (West, Westlaw through 2014 Sess.) (permitting outpatient commitment if the person’s mental illness “limits or negates his ability to make an informed decision to seek voluntarily or comply with recommended treatment”).
110. The absence of an incapacity or incompetence requirement in Kendra’s Law has been the subject of considerable controversy. See infra note 187 and accompanying text.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 177
Kendra’s Law limits the class of persons who may petition for assisted
outpatient treatment to the following: (i) persons eighteen years of age or older
with whom the subject of the petition resides; (ii) the parent, spouse, sibling or
child of the subject of the petition; (iii) the director of a hospital in which the
subject of the petition is hospitalized; (iv) the director of a charitable
organization; or (v) a qualified psychiatrist or psychologist.111 Petitions for
assisted outpatient treatment must be accompanied by an affirmation or
affidavit from a physician—other than the petitioner—stating that he or she has
personally examined the subject of the petition no more than ten days prior to
submitting the petition, recommends assisted outpatient treatment, and is able
to testify during a hearing on the petition.112
In New York, involuntary hospitalization requires a finding that the subject
of a petition for inpatient commitment presents “a substantial risk of physical
harm” to self or others.113 By contrast, Kendra’s Law permits outpatient
commitment, largely on the ground that treatment noncompliance has led to
multiple hospitalizations and without outpatient commitment, the person is
likely to decompensate, becoming a danger to himself or others. The result is
that the subject of an AOT petition can be ordered to comply with treatment,
even though at present he or she does not present a substantial risk of physical
harm to self or others. If the subject of the petition meets the criteria for AOT,
the court may order assisted outpatient treatment for up to one year.114 Thirty
days prior to the expiration of an AOT order, the petitioner may seek continued
assisted outpatient treatment for up to one year.115
Under Kendra’s Law, a court may order a person to self-administer
psychotropic drugs or accept the administration of such drugs by authorized
personnel.116 However, like most outpatient commitment statutes, Kendra’s
Law does not authorize forced administration of medication over the patient’s
objection.117 If a patient refuses to comply with any aspect of the AOT order,
and a physician determines that the patient may be in need of involuntary
111. N.Y. MENTAL HYG. LAW §§ 9.60(e)(1)(i)-(vi) (McKinney, Westlaw through 2014 Sess.).
112. Id. § 9.60(e)(3)(i).
113. N.Y. MENTAL HYG. LAW §§ 9.37(a)(1)-(2) (McKinney, Westlaw through 2014 Sess.).
114. N.Y. MENTAL HYG. LAW § 9.60(k) (McKinney, Westlaw through 2014 Sess.).
115. Id.
116. Id. § 9.60(h)(i).
117. If a person requires medication, a clinician may petition the court for an order to administer medication over his or her objection. The court will schedule a “Rivers hearing” to determine whether the person lacks the capacity to make a treatment decision. If the court concludes that the person lacks the capacity to make a treatment decision, the court must determine whether “the proposed treatment is narrowly tailored to give substantive effect to the patient’s liberty interest, taking into consideration all relevant circumstances, including the patient’s best interests, the benefits to be gained from the treatment, the adverse side effects associated with the treatment and any less intrusive alternative treatments.” Rivers v. Katz, 67 N.Y.2d 485, 497 (N.Y. 1986). The state bears the burden to establish each of these elements by clear and convincing evidence.
178 STANFORD LAW & POLICY REVIEW [Vol. 26:159
hospitalization, patients may be removed from the community and detained in a
hospital where they can be held for up to seventy-two hours to determine
whether they meet the criteria for inpatient civil commitment.118
b. Program Implementation
New York spends about $32 million dollars each year on the assisted
outpatient treatment program.119 In addition to funding the AOT Program, New
York spends $125 million dollars each year on enhanced community services
to benefit people in the AOT program as well as people who received voluntary
mental health services outside of the AOT program.120
According to the New York State Office of Mental Health, 12,421 New
Yorkers have been ordered to participate in the assisted outpatient treatment
program since the program began in November 1999.121 A large majority of
AOT cases (70%) are concentrated in New York City, and the vast majority of
AOT petitions (84%) are filed before the subject of the petition has been
discharged from the hospital.122 Nineteen percent of AOT orders are in place
for six months or less, 81% of AOT orders are in place for six months or more
and 23% of AOT orders are in place for thirty months or more.123 Most AOT
recipients are men—of whom about half are white, 21% are African American
and 30% are Hispanic. Three-quarters of AOT recipients have been diagnosed
with schizophrenia; 20% have a diagnosis of bipolar disorder,124 and a sizeable
number (43%) have a comorbid substance abuse disorder.125
118. N.Y. MENTAL HYG. LAW § 9.60(n); see also SAFE Act, supra note 7 (strengthening state gun control laws by: (i) extends Kendra’s Law for two years from its original sunset date of June 30, 2015 to June 30, 2017; (ii) extends the maximum duration of an initial AOT order from 6 months to 1 year; (iii) mandates a review by the local director of community services within 30 days prior to the expiration of an AOT order; (iv) authorizes AOT treatment order across county lines; and (v) requires a clinical assessment for an inmate committed to a state correctional facility from a psychiatric hospital prior to discharge).
119. Spending on the AOT Program includes $9.55 million on case managers, $15 million for a medication grant program, $4.4 million for prison and jail discharge managers, $2.4 million for AOT oversight programs, and $0.65 million on drug monitoring. MARTIN S. SWARTZ ET AL., NEW YORK STATE ASSISTED OUTPATIENT TREATMENT PROGRAM
EVALUATION 46 (2009), available at http://www.omh.ny.gov.
120. Id. The New York State Office of Mental Health has used those funds to increase the number of assertive community treatment teams and intensive case managers, while also developing a Single Point of Access Program (SPOA).
121. N.Y. OMH AOT REPORTS, http://bi.omh.state.ny.us/aot/characteristics (last visited Dec. 31, 2014).
122. MARTIN S. SWARTZ ET AL., NEW YORK STATE ASSISTED OUTPATIENT TREATMENT
PROGRAM EVALUATION 6 (2009), available at http://www.omh.ny.gov.
123. N.Y. OMH AOT REPORTS, http://bi.omh.ny.gov/aot/statistics?p=time-in-aot (last visited June 20, 2014).
124. N.Y. OMH AOT REPORTS, http://bi.omh.ny.gov/aot/characteristics?p=diagnosis-diagnosis (last visited June 20, 2014).
125. N.Y. OMH AOT REPORTS, http://bi.omh.ny.gov/aot/characteristics?p=diagnosis-alcohol-sub-abuse (last visited June 20, 2014).
2015] INVOLUNTARY OUTPATIENT COMMITMENT 179
2. California
Like Kendra’s Law, California’s outpatient commitment statute—known as
Laura’s Law—was passed in response to an act of violence. The State
Legislature named Laura’s Law in memory of Laura Wilcox, a nineteen-year-
old college sophomore, who was volunteering at a mental health clinic in
Nevada City, California, when Scott Thorpe entered the clinic and opened
fire.126 Thorpe shot and killed Wilcox and another volunteer and then drove to
a restaurant where he shot and killed the manager.127 Scott Thorpe had a long
history of mental illness, including a delusional belief that the FBI was trying to
poison him and compel him to see an incompetent psychiatrist.128
Laura’s Law was modeled on Kendra’s Law, and most of the criteria for
issuing an assisted outpatient treatment order are the same. Under Laura’s Law,
a court may order a person to comply with an assisted outpatient treatment
order if it finds by clear and convincing evidence that the person is: (i) eighteen
years of age or older; (ii) suffering from a mental illness; and (iii) unlikely to
survive safely in the community without supervision.129 The petitioner must
also establish that a history of treatment noncompliance has resulted either in
(i) hospitalization or treatment at least twice within the last thirty-six months or
(ii) one or more acts of serious and violent behavior toward self or others (or
threats or attempts at such acts) within the last forty-eight months.130 Like
Kendra’s Law, the subject of a court order must be likely to benefit from
assisted outpatient treatment, and the petitioner must establish that treatment is
necessary to prevent a relapse or deterioration.131
Laura’s Law also includes several requirements that are far more restrictive
than Kendra’s Law. To qualify for an assisted outpatient treatment order under
Laura’s Law, the petitioner must establish that the person’s condition is
“substantially deteriorating,”132 while Kendra’s Law only requires evidence
that the person is likely to deteriorate without outpatient treatment.133
Similarly, Kendra’s Law requires evidence that the person is unlikely to
participate in outpatient treatment voluntarily,134 however, Laura’s Law
requires evidence that the person has actually been given the opportunity to
participate in outpatient treatment program and failed to do so.135 In contrast to
126. Logan Jenkins, As County Considers Laura’s Law A Mother Reflects, UNION-TRIB. SAN DIEGO, Mar. 8, 2013, http://www.utsandiego.com/news/2013/mar/08/laura-law-county-nevada.
127. Editorial, Why It’s Called Laura’s Law, S.F. CHRON., May 1, 2002, at A20.
128. Id.
129. CAL. WELF. & INST. CODE §§ 5346(a)(1)-(3) (West, Westlaw through 2014 Sess.).
130. Id. §§ 5346(a)(4)(A), (B).
131. Id. §§ 5346(a)(8)-(9).
132. Id. § 5346(a)(6).
133. N.Y. MENTAL HYG. LAW § 9.60(c)(6) (McKinney, Westlaw through 2014 Sess.).
134. Id. § 9.60(c)(5).
135. CAL. WELF. & INST. CODE § 5346(a)(6).
180 STANFORD LAW & POLICY REVIEW [Vol. 26:159
Kendra’s Law, friends, family members, and clinicians may ask the county
mental health department to petition a court for outpatient treatment, but only
the county mental health director, or a designee, may file the petition.136
Laura’s Law was passed in 2002 and signed into law by Governor Gray
Davis. Yet the statute only applies in counties where the board of supervisors
authorizes the implementation of Laura’s Law.137 Laura’s Law imposes a long
list of requirements on any county that wants to create an assisted outpatient
treatment program. In order to create an AOT program, counties must provide
mobile mental health teams with high staff-to-client ratios of no more than ten
clients per provider, develop plans for outreach to families with severely
mentally ill adults, and develop a plan for housing clients.138 Counties must
provide services for seriously mentally ill adults who are under twenty-five
years old and at risk of homelessness, as well as culturally appropriate services
for women from diverse ethnic backgrounds, including supportive housing that
accepts children, substance abuse programs that address gender specific
trauma, and vocational programs that are sensitive to the needs of women.139
Further, any county that provides assisted outpatient treatment under
Laura’s Law must provide the same services to persons who seek them
voluntarily.140 Counties must also show that voluntary mental health services
will not be reduced by creating an assisted outpatient treatment program.141
Counties can use state mental health funds to implement Laura’s Law.142
However, in contrast to New York, where the passage of Kendra’s Law
coincided with an annual infusion of $157 million dollars into the state mental
health budget, California has not made new funds available to implement
Laura’s Law, and the mental health budget in California is substantially
underfunded.143 Of California’s fifty-eight counties, thus far, only Nevada
County (where Laura Wilcox lived), San Francisco County, Orange County,
and Los Angeles County have implemented assisted outpatient treatment
programs.144
136. Id. § 5346(b)(1).
137. CAL. WELF. & INST. CODE § 5348(a) (West, Westlaw through 2014 Sess.).
138. Id. §§ 5348(a)(1), (2)(B).
139. Id. §§ 5348(a)(2)(H)-(I).
140. Id. § 5348(b).
141. CAL. WELF. & INST. CODE § 5349 (West, Westlaw through 2014 Sess.).
142. Id.
143. See Paul S. Appelbaum, Ambivalence Codified: California’s New Outpatient Commitment Statute, 54 PSYCHIATRIC SERVICES 26, 27 (2003) (characterizing Laura’s Law as “a sham”).
144. Jill Cowan & Paloma Esquivel, Addressing the Mentally Ill: O.C. is the First Large County in California to Implement Laura’s Law, L.A. TIMES, May 14, 2014, at AA.1; Abby Sewell, L.A. County to Expand Laura’s Law Mental-Illness Treatment Program, L.A. TIMES, July 15, 2014, http://www.latimes.com/local/countygovernment/la-me-lauras-law-20140716-story.html.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 181
3. North Carolina
Long before New York and California established their outpatient
commitment programs, North Carolina became the first state to enact a
preventive outpatient commitment statute in 1983.145 Under the North Carolina
statute, any person who knows an individual to be mentally ill and a danger to
himself or others and in need of treatment may petition a court for an order to
take the person into custody for an examination by a psychologist or a
physician.146 If the person meets the criteria for outpatient commitment, the
examiner may recommend outpatient commitment and notify the court.147
Unless the person also meets the criteria for inpatient commitment, the person
must be released from custody and returned to his or her home.148
A North Carolina court may order outpatient treatment if it finds by clear,
cogent, and convincing evidence that the person is (i) mentally ill; (ii) capable
of surviving safely in the community with supervision; and (iii) needs treatment
to prevent further disability or deterioration which would “predictably result in
dangerousness.”149 Unlike most states, the North Carolina outpatient
commitment statute contains an incapacity requirement. An outpatient
commitment order under North Carolina law requires evidence that the
person’s mental illness “limits or negates his ability to make an informed
decision to seek voluntarily or comply with recommended treatment.”150 If a
person refuses to comply with an outpatient treatment plan, providers are
required to make “all reasonable effort” to encourage cooperation.151 The
person may be taken into custody to determine whether he or she meets the
criteria for inpatient commitment.152
North Carolina does not permit medication by force without evidence of an
immediate danger to self or others.153 However, patients might believe they
will be forced to take their medications nonetheless. In a study of 306
outpatients in North Carolina, 82.7% of respondents believed that they were
required to take their medications, even though North Carolina law does not
145. Erica F. King, Outpatient Civil Commitment in North Carolina: Constitutional and Policy Concerns, 58 LAW & CONTEMP. PROBS. 251 (1995).
146. See N.C. GEN. STAT. ANN. § 122C-261(a) (West, Westlaw through 2014 Sess.) (requiring that once the person has been admitted to a hospital, an examination must take place within twenty-four hours); N.C. GEN. STAT. ANN. § 122C-267(a) (West, Westlaw through 2014 Sess.) (requiring that a hearing must take place no more than ten days after the person was initially taken into custody).
147. N.C. GEN. STAT. ANN. § 122C-264(a) (West, Westlaw through 2014 Sess.).
148. N.C. GEN. STAT. ANN. § 122C-263(d) (West, Westlaw through 2014 Sess.).
149. Id.
150. N.C. GEN. STAT. ANN. § 122C-271(a)(1) (West, Westlaw through 2014 Sess.).
151. N.C. GEN. STAT. ANN. § 122C-273(a)(1) (West, Westlaw through 2014 Sess.).
152. N.C. GEN. STAT. ANN. § 122C-273(a)(2).
153. N.C. GEN. STAT. ANN. § 122C-273(a)(3).
182 STANFORD LAW & POLICY REVIEW [Vol. 26:159
authorize providers to force medication over an outpatient’s objection.154 As
the authors note, outpatients may have believed that they can be forced to
comply with medication since they may have observed or experienced forced
medication in an inpatient setting.155
C. Empirical Research on Effectiveness
The moral limits on our ability to intervene in the lives of people with
mental illnesses before they harm themselves or others are the primary subject
matter of this Article. Parts IV and V explore that terrain. Before turning to
those questions, I shall briefly summarize debates on the effectiveness of
outpatient commitment and the constitutional legitimacy of these laws in Parts
II.C and II.D. Although the major contribution of this Article is not to resolve
either of those debates, understanding each of them frames the normative
discussion in which I engage.
After two generations of research on involuntary outpatient commitment,
the empirical evidence on effectiveness is mixed at best.156 However, recent
studies suggest that participation in a well-funded outpatient commitment
program may be associated with significant benefits for people with mental
illnesses, including fewer hospitalizations, shorter hospital stays, and a greater
likelihood of receiving appropriate psychotropic medications. I am willing to
assume that outpatient commitment laws are associated with these benefits.
Nonetheless, in Part IV, I argue that the special respect we owe to persons as
moral agents prevents us from trading their liberties for mere gains in utility.
1. Randomized Controlled Trials
There have been three randomized controlled trials of outpatient
commitment. In the Duke Mental Health Study, 331 involuntarily hospitalized
patients, who were awaiting discharge on an outpatient commitment order in
North Carolina, were randomly assigned to receive outpatient commitment or
have their commitment order dismissed.157 Patients in both groups were given
154. Randy Borum et al., Consumer Perceptions of Involuntary Outpatient Commitment, 50 PSYCHIATRIC SERVICES 1489 (1999).
155. Id. at 1490. Critics of outpatient commitment argue that failing to educate patients on the consequences of treatment noncompliance violates basic principles of informed consent. See, e.g., BRUCE J. WINICK, CIVIL COMMITMENT, A THERAPEUTIC JURISPRUDENCE
MODEL, 254 (2005) (characterizing the approach as “subterfuge.”); Michael A. Hoge & Elizabeth Grottole, The Case Against Outpatient Commitment, 28 J. AM. ACAD. PSYCHIATRY
& L. 165, 167 (2000).
156. See M. SUSAN RIDGELY, ET AL., RAND HEALTH: RAND INSTITUTE FOR CIVIL
JUSTICE, THE EFFECTIVENESS OF INVOLUNTARY OUTPATIENT TREATMENT: EMPIRICAL
EVIDENCE AND THE EXPERIENCE OF EIGHT STATES 17-22 (2001) (summarizing early research on outpatient commitment).
157. Marvin S. Swartz et al., A Randomized Controlled Trial of Outpatient Commitment in North Carolina, 52 PSYCHIATRIC SERVICES 325 (2001).
2015] INVOLUNTARY OUTPATIENT COMMITMENT 183
case managers and outpatient treatment at one of four area mental health
programs. Researchers found that when the duration of the outpatient
commitment order was not taken into account, the outcomes of interest—
hospital readmissions, hospital days, and violent behavior—did not differ
significantly between the control group and the experimental group.158 The
study also found that patients who had extended periods of outpatient
commitment (180 days or more) and intensive outpatient treatment (seven
service days per month on average) had 57% fewer hospital readmissions and
twenty fewer hospital days per month when compared to the control group.159
The study found a similar relationship between outpatient commitment and
violence. Outpatient commitment did not reduce the risk of violence among
patients with fewer than three service visits per month. However, patients with
sustained outpatient commitment orders of 180 days or more and three or more
service visits per month were significantly less likely to be violent (twenty-four
versus forty-eight percent).160 While a court order alone cannot improve patient
outcomes, the researchers hypothesized that court orders might have worked by
incentivizing providers to prioritize treatment and motivating patients to
comply.161
The methods employed by the Duke Mental Health Study had some
limitations. For example, the researchers used a stringent adherence protocol in
which sheriffs obtained pick-up orders for patients who missed three or more
consecutive appointments and clinicians promptly rescheduled missed
appointments.162 However, enforcement of outpatient commitment orders may
be less rigorous in practice, thereby limiting the generalizability of their
findings to non-clinical settings.163
The second randomized controlled trial of outpatient commitment took
place at Bellevue Hospital in New York.164 Seventy-eight patients were
randomly assigned to receive assisted outpatient treatment, and sixty-four
patients were assigned to a control group in which they did not receive court
158. Id. at 327.
159. Id.
160. Id.; see also Jeffrey W. Swanson et al., Involuntary Out-patient Commitment and Reduction in Violent Behaviour in Persons with Severe Mental Illness, 176 BRIT. J. PSYCHIATRY 324 (2000); Swartz et al., supra note 158, at 326.
161. Id. at 329 n.157.
162. Virginia Aldigé Hiday et al., Impact of Outpatient Commitment on Victimization of People with Severe Mental Illness, 159 AM. J. PSYCHIATRY 1403, 1405 (2002).
163. Further, although patients were randomly assigned to the outpatient commitment group or the control group, the amount of time on outpatient commitment was not randomly assigned. Patients who no longer met the criteria for outpatient commitment did not have their orders renewed. However, since patients with a greater risk of poor outcomes were more likely to have their outpatient commitment orders renewed, any selection effects would have reduced the likelihood of finding a positive effect for outpatient commitment. See also, George Szmukler & Matthew Hotopf, Effectiveness of Involuntary Outpatient Commitment, 158 AM. J. PSYCHIATRY 653 (2001) (offering a critique of the Duke Mental Health Study).
164. Henry J. Steadman et al., Assessing the New York City Involuntary Outpatient Commitment Pilot Program, 52 PSYCHIATRIC SERVICES 330 (2001).
184 STANFORD LAW & POLICY REVIEW [Vol. 26:159
ordered treatment.165 Both groups received case management, psychiatric
aftercare, and a residential placement. If a patient was in the assisted outpatient
treatment group, the treatment plan was formalized by a court order. In contrast
to the Duke Study, after a twelve month follow-up, the Bellevue Study found
no statistically significant differences between the AOT group and the non-
AOT group across all outcomes, namely rehospitalization rates, hospital days,
arrest rates, quality of life, and psychiatric symptoms.166 Nor did service
providers note significant differences in treatment adherence between the two
groups.167 Like the Duke Study, however, the methods used in the Bellevue
Study had several disadvantages. For example, AOT orders were not
consistently enforced in the Bellevue Study which might have diminished some
of the difference between the AOT group and the non-AOT group.168
Even if outpatient commitment orders were effective, a systematic review
by the Cochrane Collaboration highlighted a further concern: the numbers
needed to treat in order to produce such an effect might be unacceptably
high.169 The Cochrane Collaboration conducted a meta-analysis of randomized
controlled trials of outpatient commitment for people with mental disorders. At
the time, there were only two—the Duke Mental Health Study and the Bellevue
Study. The authors concluded that outpatient commitment orders were not an
effective alternative to voluntary treatment since courts would have to issue
eighty-five outpatient commitment orders to prevent one hospital readmission,
twenty-seven outpatient commitment orders to prevent one episode of
homelessness, and 238 outpatient commitment orders to prevent one arrest.170
Findings from a third trial of outpatient commitment in England and Wales
have reignited the debate over the effectiveness of outpatient commitment
orders in the United States.171 Participants in the Oxford Community Treatment
Order Evaluation Trial (OCTET) were randomly assigned to an experimental
group, which included a community treatment order (CTO) or treatment as
165. Id. at 331.
166. Id. at 332-33.
167. Id. at 330.
168. Id. at 335; see also Howard Telson, Outpatient Commitment in New York: From Pilot Program to State Law, 11 GEO. MASON U. C.R. L.J. 41 (2000) (offering a detailed discussion and criticism of the pilot program).
169. STEVE R. KISLEY, LESLIE A. CAMPBELL & NEIL J. PRESTON, THE COCHRANE
COLLABORATION, COMPULSORY COMMUNITY AND INVOLUNTARY OUTPATIENT TREATMENT
FOR PEOPLE WITH SEVERE MENTAL DISORDERS 19 (2011), Cochrane Database of Systematic Reviews 2011, Issue 2, Article No. CD004408, DOI: 10.1002/14651858.CD004408.pub3.
170. Id. at 2.
171. See, e.g., Jeffrey Swanson & Marvin Swartz, Why the Evidence for Outpatient Commitment Is Good Enough, 65 PSYCHIATRIC SERVICES 808 (2014) (commenting on OCTET); TREATMENT ADVOC. CENTER, NO RELEVANCE TO ASSISTED OUTPATIENT
TREATMENT (AOT) IN THE OCTET STUDY OF ENGLISH COMPULSORY TREATMENT, http://www.treatmentadvocacycenter.org/storage/documents/Research/may2013-octet-study.pdf.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 185
usual (i.e., a “leave of absence” under Section 17 of the Mental Health Act).172
A Section 17 leave of absence allows patients to leave the hospital for days or
weeks, with less monitoring, while remaining subject to recall. Community
treatment orders resemble outpatient commitment orders in the U.S. insofar as
CTOs can direct patients to take medications outside of the hospital and subject
patients to intensive monitoring, but they do not authorize clinicians to
administer medications by force.173 The primary outcome of interest was
whether patients were readmitted to a hospital during the twelve-month study
period. The study found no statistically significant differences between patients
on community treatment orders and patients on a leave of absence, with respect
to hospitalization rates, time to readmission, and hospital days over the twelve-
month study period.174
The results from the OCTET evaluation have also been challenged on
methodological grounds. Critics argue that OCTET was not adequately
designed to determine whether outpatient commitment orders are more
effective than voluntary treatment for people with mental illnesses.175 The
control condition in OCTET was a Section 17 leave of absence, not voluntary
treatment. A Section 17 leave in the United Kingdom involves supervised
community treatment, including clinical monitoring and the power to recall
patients to the hospital, therefore it is not surprising that the study did not find
statistically significant differences between the control condition and
community treatment orders.176
2. Observational and Quasi-Experimental Designs
Supporters of outpatient commitment argue that the findings from a 2009
evaluation of the assisted outpatient treatment program in New York are
sufficient to demonstrate that court ordered treatment is probably more
effective for hard-to-reach populations than voluntary services alone. The
study, led by Marvin Swartz and Jeffrey Swanson, compared people who
received assertive community treatment with a court order to people who
received assertive community treatment without a court order.177 Among
172. Tom Burns et al., Community Treatment Orders for Patients with Psychosis (OCTET): A Randomised Controlled Trial, 381 LANCET 1627 (2013).
173. Id. at 1629.
174. Id. at 1631.
175. See, e.g., Swanson & Swartz, supra note 171 (discussing a critique of the methods used in OCTET).
176. Id. at 810.
177. MARVIN S. SWARTZ ET AL., NEW YORK STATE ASSISTED OUTPATIENT TREATMENT
PROGRAM EVALUATION (2009), available at http://www.omh.ny.gov; Assertive Community Treatment (ACT) Teams provide treatment and intensive case management for patients in the community. ACT Teams are interdisciplinary teams of ten to twelve professionals, including a case manager, a psychiatrist, a social worker, nurses, vocational, and substance abuse counselors and peer counselors. DEP’T OF HEALTH AND HUMAN SERVICES: U.S. PUBLIC
HEALTH SERV., MENTAL HEALTH: A REPORT OF THE SURGEON GENERAL 286-87 (1999).
186 STANFORD LAW & POLICY REVIEW [Vol. 26:159
people who received assertive community treatment in the AOT program for
twelve months or more, hospitalizations were reduced by about one-half—from
seventy-four percent at baseline to thirty-six percent.178 By contrast, among
people who only participated in assertive community treatment without a court
order, psychiatric hospitalizations were reduced by about one-third—from six-
ty-three percent at baseline to fifty-eight percent after twelve months.179
Swartz and Swanson also asked whether AOT recipients were less likely to
be hospitalized and more likely to receive appropriate medications when
compared with their own experience before participating in the assisted
outpatient treatment program. Researchers examined Medicaid data, hospital
records and case manager reports for 2839 people who were enrolled in the
AOT program and Medicaid at any time between January 1999 and March
2007.180 The study found that participation in the AOT program was associated
with a substantial reduction in the probability of a psychiatric hospitalization
and the number of days hospitalized. The probability of a psychiatric
hospitalization for the average AOT recipient dropped from fourteen percent
prior to AOT to eleven percent during the first six months of AOT, and to nine
percent during the next seven to twelve months of AOT.181 The study reported
similar reductions in the average duration of a hospital stay—from eighteen
days before AOT to eleven hospital days during the first six months of AOT
and ten hospital days during the next seven to twelve months of AOT.182 By
examining Medicaid claims data and the dates when prescriptions were filled,
researchers were also able to calculate the proportion of days in each month
that AOT recipients would have had an adequate supply of medication. The
percentage of AOT recipients who had an adequate supply of medication
during eighty percent of the month or more increased from thirty-five percent
prior to AOT to forty-four percent during the first six months of AOT and then
to fifty percent during the next secent to twelve months of AOT.183
The same 2009 study by Swartz and Swanson found that people who
participated in the AOT program were less likely to be arrested than their
counterparts in voluntary treatment.184 If AOT orders were kept in place for six
months or more, AOT recipients were more likely to have the right medications
and less likely to be hospitalized, even after the court order ended.185
178. SWARTZ ET AL., supra note 177, at 27.
179. Id.
180. Id. at 28.
181. Id. at 29.
182. Id. at 30.
183. Id. at 31.
184. Id. at 23; see also Allison R. Gilbert et al., Reductions in Arrest Under Assisted Outpatient Treatment in New York, 61 PSYCHIATRIC SERVICES 996 (2010).
185. SWARTZ ET AL., supra note 177, at 39-41; see also Richard A. Van Dorn, Continuing Medication and Hospitalization Outcomes After Assisted Outpatient Treatment in New York, 61 PSYCHIATRIC SERVICES 982 (2010).
2015] INVOLUNTARY OUTPATIENT COMMITMENT 187
Yet as the authors concede, the study findings are not without
limitations.186 The study used a matching design to compare AOT recipients to
people who received treatment voluntarily. However, since a matching design
would only allow the authors to control for observed, but not unobserved
covariates, the findings from a matched design are less definitive than the
results of a careful randomized controlled trial. Nor are the findings from New
York’s assisted outpatient treatment program clearly generalizable to other
states like California, where AOT programs are underfunded.
Debate about whether outpatient commitment laws are an effective
solution to the problem of revolving door hospitalizations for people with
mental illnesses will undoubtedly continue. For the purposes of this Article, I
am willing to bracket this uncertainty while I discuss the moral questions these
laws raise. Before turning to that discussion, I will outline the constitutional
framework in which these laws operate.
D. Constitutional Challenges
The legal controversy surrounding outpatient commitment has focused on
the substantive and procedural due process rights of people with mental
illnesses.187 Although the U.S. Supreme Court has yet to consider the
constitutional questions presented by outpatient commitment, the Court has
recognized a liberty interest in refusing unwanted medical treatment188 as well
as a liberty interest in refusing unwanted antipsychotic medication,189 both
arising from the Due Process Clause.
The leading case on the constitutionality of outpatient commitment came
from the New York Court of Appeals in In re K.L.190 K.L. alleged that
186. SWARTZ ET AL., supra note 177, at 51.
187. See, e.g., Jennifer Gutterman, Waging a War on Drugs: Administering a Lethal Dose to Kendra’s Law, 68 FORDHAM L. REV. 2401 (1999); Erin O’Connor, Is Kendra’s Law a Keeper? How Kendra’s Law Erodes Fundamentally the Rights of the Mentally Ill, 11 J.L. & POL’Y 313 (2002); Susan Stefan, Preventive Outpatient Commitment: The Concept and its Pitfalls, 11 MENTAL & PHYSICAL DISABILITY L. REP. 288 (1987). But see Jeffrey Geller & Jonathan A. Stanley, Settling the Doubts About the Constitutionality of Outpatient Commitment, 31 NEW. ENG. J. ON CRIM & CIV. CONFINEMENT 127 (2005); Illisa Watnick, A Constitutional Analysis of Kendra’s Law: New York’s Solution for the Treatment of the Chronically Mentally Ill, 149 U. PA. L. REV. 1181 (2001) (arguing that Kendra’s Law does not violate substantive due process).
188. Cruzan ex rel. Cruzan v. Director, Missouri Dep’t of Health, 497 U.S. 261, 279 n.7 (1990) (“Although many state courts have held that a right to refuse treatment is encompassed by a generalized constitutional right of privacy, we have never so held. We believe this issue is more properly analyzed in terms of a Fourteenth Amendment liberty interest.”).
189. Mills v. Rogers, 457 U.S. 291, 299 (1982) (asserting that “the Constitution recognizes a liberty interest in avoiding the unwanted administration of antipsychotic drugs”); Washington v. Harper, 494 U.S. 210 (1990); Riggins v. Nevada, 504 U.S. 127 (1992).
190. In re K.L., 806 N.E.2d 480 (N.Y. 2004).
188 STANFORD LAW & POLICY REVIEW [Vol. 26:159
Kendra’s Law violates due process because it does not require a judicial
determination of incapacity before a court may order a person to comply with
an assisted outpatient treatment plan.191 In pressing his claims, K.L. relied upon
Rivers v. Katz, where the New York Court of Appeals discussed the
circumstances in which a hospital can administer antipsychotic medication to
an involuntarily committed patient against his or her will.192 Drawing upon
familiar language from Justice Cardozo in Schloendorff v. Society of New York
Hospital, the Court of Appeals in Rivers declared: “[i]t is a firmly established
principle of the common law of New York that every individual ‘of adult years
and sound mind has a right to determine what shall be done with his [or her]
own body.’”193 Moreover:
In our system of a free government where notions of individual autonomy
and free choice are cherished, it is the individual who must have the final say
in respect to decisions regarding his medical treatment in order to insure that
the greatest possible protection is accorded his autonomy . . . . This right
extends equally to mentally ill persons who are not to be treated as persons of
lesser status or dignity because of their illness.194
In Rivers, the New York Court of Appeals observed that the right of a
competent person to refuse treatment safeguards is an individual interest in
autonomy, an interest which is no less important simply because the rightholder
has been diagnosed with a mental illness. The Court of Appeals affirmed a
fundamental right to refuse treatment in Rivers, but noted that the right to
refuse treatment is not absolute and instead may yield to compelling state
interests.195 Importantly, however, the Rivers Court held that while the State
has the parens patriae power to provide treatment to people who are unable to
care for themselves due to mental illness, “[f]or the State to invoke this interest,
‘the individual himself must be incapable of making a competent decision
concerning treatment on his own. Otherwise, the very justification for the
state’s purposed exercise of its parens patriae power—its citizen’s inability to
care for himself * * * would be missing.’”196 Therefore, the Court reasoned,
when a patient refuses to consent to antipsychotic drugs, there must be a
judicial hearing to determine whether the patient has the capacity to make a
reasoned treatment decision before a hospital may administer drugs pursuant to
the parens patriae power.197
The Court of Appeals rejected the Rivers analogy in In re K.L. and
declined to strike Kendra’s Law on substantive due process grounds. In contrast
to the circumstances at issue in Rivers, AOT patients cannot be forced to accept
191. Id. at 483-84.
192. Rivers v. Katz, 67 N.Y.2d 485 (1986); In re K.L., 806 N.E.2d at 484.
193. Id. at 493 (citing Schloendorff v. Society of N.Y. Hosp, 211 N.Y. 125, 129 (1914)).
194. Id. at 493.
195. Id. at 495.
196. Id. at 496 (citing Rogers v. Okin, 634 F.2d 650, 657 (1980)).
197. Id. at 498.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 189
antipsychotic medications. Since the force of an assisted outpatient treatment
order “lies solely in the compulsion generally felt by law-abiding citizens to
comply with court directions,” the Court of Appeals characterized the restraint
on the right to refuse treatment as “minimal.”198 Moreover, the Court
concluded that assisted outpatient treatment orders are an appropriate exercise
of the State’s parens patriae power since an AOT order requires several
findings, including evidence that the person is unlikely to survive safely in the
community without supervision; a history of treatment noncompliance; and
evidence that the person is unlikely to comply with treatment voluntarily.199
The Court also found ample authority for Kendra’s Law in the State’s police
power to “protect the community from the dangerous tendencies of some who
are mentally ill.”200
In the same way, the Court of Appeals summarily rejected K.L.’s
procedural due process claim. K.L. alleged that failing to provide notice and a
hearing before a noncompliant patient may be removed from the community
and detained in a hospital violates procedural due process.201 The Court of
Appeals agreed that involuntary hospitalization constitutes a substantial
deprivation of liberty. Nonetheless, the Court concluded that the State’s interest
in removing patients, previously found to be at risk of relapse or deterioration,
outweighs the patient’s liberty interests in freedom from an unwanted
hospitalization. Nor was the Court persuaded that the Fourth Amendment
requires probable cause before a physician can remove a noncompliant patient.
Since the statute directs physicians to use their “clinical judgment,” it
“necessarily contemplates that the [removal] determination will be based on the
physician’s reasonable belief that the patient is in need of such care.”202
The Court of Appeals was equally unpersuaded by the plaintiff’s equal
protection arguments. Without a judicial determination of incapacity, the
plaintiffs maintained that Kendra’s Law violates equal protection since persons
under guardianship and involuntarily committed psychiatric patients are
entitled to an incapacity hearing before they can be medicated against their
will.203 The Court of Appeals simply reiterated that an assisted outpatient
treatment order does not authorize the use of force to administer medication,
and as a result, “[t]he statute thus in no way treats similarly situated persons
differently.”204
Whatever one might think about the plaintiff’s equal protection or Fourth
Amendment claims, the reasoning employed by the Court of Appeals stands on
shaky ground. The Court employed an unduly narrow interpretation of the
198. Id. at 485.
199. Id. at 498.
200. Id. at 485 (internal quotations omitted).
201. Id. at 486.
202. Id. at 487-88.
203. Id. at 486.
204. Id.
190 STANFORD LAW & POLICY REVIEW [Vol. 26:159
rights announced in Schloendorff and Rivers. Undoubtedly, the right to refuse
treatment comprehends not only the right to refuse antipsychotic medications,
but also the right to refuse outpatient treatment; individual or group therapy;
day or partial day programming; supervised living arrangements; substance
abuse counseling; and home visits by assertive community treatment teams—
all hallmarks of the typical AOT order. Second, the Court of Appeals alluded to
the “dangerous tendencies” of the mentally ill as a justification for state action
through police power. Yet large epidemiological studies have consistently
shown that most people with mental illnesses are no more dangerous than
anyone else, and more importantly, social science has progressed beyond
thinking of dangerousness as a property of the person.205 Current approaches to
risk assessment regard the risk of violence as an interaction between clinical
characteristics, demographic factors and the person’s environment.206
Finally, in contrast to Rivers, the Court of Appeals employed a dangerously
expansive interpretation of the requirements for state action under the parens
patriae power. In Rivers, the Court of Appeals held that for the State to compel
treatment through its parens patriae authority, it must be the case that the
individual is incapable of making a competent treatment decision on his own.
Otherwise the primary justification for invoking parens patriae authority
“would be missing.”207 Yet in In re K.L., the Court of Appeals reasoned that
assisted outpatient treatment orders are an appropriate exercise of the State’s
parens patriae power—even without evidence of incapacity—since an AOT
order requires several other findings, including: evidence that the person is
unlikely to survive safely without treatment; the person has a history of
treatment noncompliance; the person needs assisted outpatient treatment; and
the person would benefit from assisted outpatient treatment.208 The same logic
would justify intrusions into the decisions of cancer patients and diabetics
without a clear stopping point. Yet we do not require these patients to accept
medical treatment simply because the person needs treatment or would benefit
from treatment. Nor should we. As the Court of Appeals observed in Rivers, in
our liberal democracy, where individual interests in privacy and personal
autonomy in treatment decisions are protected, “the right of a competent adult
to refuse medical treatment must be honored, even though the recommended
treatment may be beneficial or even necessary to preserve the patient’s life.”209
205. See, e.g., Henry J. Steadman, From Dangerousness to Risk Assessment of Community Violence: Taking Stock at the Turn of the Century, 28 J AM. ACAD. PSYCHIATRY
& L. 265 (2000).
206. MARC J. ACKERMAN, ESSENTIALS OF FORENSIC PSYCHOLOGICAL ASSESSMENT 160 (2010); see also Sue E. Estroff & Catherine Zimmer, Social Networks, Social Support and Violence, in VIOLENCE AND MENTAL DISORDER: DEVELOPMENTS IN RISK ASSESSMENT 259, 271 (John Monahan & Henry J. Steadman eds., 1994) (observing that “violence seldom happens unilaterally” and is instead influenced by complex interpersonal histories, mutual hostility and longstanding family disputes.).
207. Rivers v. Katz, 67 N.Y.2d 485, 496 (N.Y. 1986).
208. In re K.L., 806 N.E.2d 480, 485-86 (N.Y. 2004).
209. Rivers, 67 N.Y.2d at 493.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 191
II. MORAL JUSTIFICATIONS AND PUBLIC POLICY RATIONALES
In In re K.L., the New York Court of Appeals mapped the constitutional
boundaries of government authority to compel outpatient commitment without
passing judgment on the wisdom of these laws or the moral claims that citizens
might have. In Part III, I examine and refute common arguments for outpatient
commitment based on harm to others, harm to self, and impaired insight.
A. Harm to Others
Supporters of outpatient commitment have argued that in many cases of
actual or threatened violence by a person with a mental illness, the perpetrator
was either not being treated for his or her mental disorder or not taking
prescribed medications.210 Moreover, most of these incidents, many of them
homicides, could have been prevented if there were laws in place to insist that
people with mental illnesses participate in treatment whether they want to or
not.211 What should we make of these claims? Part III.A reviews empirical
evidence on the relationship between mental illness and violence. Although
support for outpatient commitment stems from high profile acts of violence
committed by people with mental illnesses, most violent crimes are not
committed by people with mental illnesses, and most people with mental
illnesses are no more violent than anyone else. A second argument concedes
that most people with mental illnesses are no more dangerous than members of
the general population, but insists that a subgroup of people with mental
illnesses are more dangerous.212 To that end, involuntary outpatient
commitment laws are necessary to prevent tragedies, like the death of Kendra
Webdale, from happening again.213 Call this the preventable tragedies
argument.214
As the remainder of Part III.A asserts, the problem with this argument is
twofold. First, using involuntary outpatient commitment to prevent tragedies,
such as the Webdale incident, from happening again presumes that we have a
reasonably reliable way to identify people with mental illnesses who are likely
to be violent and distinguish them from those who are not. However, clinical
predictions of violence are only slightly better than chance. Second, even if we
210. Ken Kress, An Argument for Assisted Outpatient Treatment for People with Serious Mental Illness Illustrated with Reference to a Proposed Statute for Iowa, 85 IOWA L. REV. 1269, 1283 (2000).
211. Id.
212. See, e.g., E. Fuller Torrey, Violent Behavior by Individuals with Serious Mental Illness, in INSIGHT AND PSYCHOSIS 269 (Xavier Amador & Anthony S. David eds., 1998).
213. Ilissa L. Watnick, Comment, A Constitutional Analysis of Kendra’s Law: New York’s Solution for Treatment of the Chronically Mentally Ill, 149 U. PA. L. REV. 1181, 1183 (2001).
214. See, e.g., The Preventable Tragedies Database, TREATMENT ADVOC. CENTER, http://www.treatmentadvocacycenter.org/home-page/71-featured-articles/2014-how-google-and-the-media-shape-our-preventable-tragedies-database.
192 STANFORD LAW & POLICY REVIEW [Vol. 26:159
were able to identify people with mental illnesses who are likely to be violent,
we can identify statistically significant associations between violence and any
number of risk factors—age and violence, gender and violence, income and
violence, educational attainment, and violence. However, as a general matter,
the other-regarding harms we aim to prevent through outpatient commitment
are addressed retrospectively, through the criminal justice system. Even if we
were able to identify a subgroup of the population as very likely to engage in
violence, courts will not impose limits on their freedom in order to prevent the
very serious crimes that they are likely to commit. Proponents of the
preventable tragedies argument will need to explain why people with mental
illnesses should be treated differently.
1. Violence and Mental Illness
a. Community Surveys
Several epidemiological studies have found a modest association between
mental illness and violence. In a seminal study on violence and mental disorder,
Jeffrey Swanson and colleagues analyzed data drawn from the Epidemiological
Catchment Area Study conducted by the National Institute of Mental Health.215
As part of the study, 10,000 randomly chosen adults were interviewed to
establish the prevalence of mental disorder. Study participants were also asked
to self-report violent behaviors during the past year (e.g., injuring a spouse or
partner, getting into physical fights, or using a weapon). The study found that
schizophrenia, bipolar disorder, and major depression were associated with a
fourfold increase in the odds of violence within one year, after controlling for
sociodemographic variables such as age, gender, socioeconomic status,
ethnicity, and race (odds ratio = 3.94).216 However, the study also found that
substance abuse was associated with a far greater risk of violence (odds ratio =
13.67).217 To put these numbers in perspective, the authors also estimated the
attributable risk of violence associated with mental disorder. Since serious
mental illnesses are rare, people with a diagnosis of mental disorder alone only
accounted for about four to five percent of the total violence in the population
over the course of one year.218 By contrast, since violence was more common
among drug and alcohol abusers, and since there were more substance abusers
in the community, the attributable risk of violence among substance abusers
was considerably higher, on the order of twenty-seven percent.219
215. Jeffrey W. Swanson, Mental Disorder, Substance Abuse, and Community Violence: An Epidemiological Approach, in VIOLENCE AND MENTAL DISORDER 101 (John Monahan & Henry J. Steadman eds., 1994).
216. Id. at 130.
217. Id.
218. Id. at 118.
219. Id.; see also Heather L. Stuart & Julio E. Arboleda-Flórez, A Public Health Perspective on Violent Offenses Among Persons with Mental Illness, 52 PSYCHIATRIC
2015] INVOLUNTARY OUTPATIENT COMMITMENT 193
Findings from the landmark MacArthur Violence Risk Assessment Study
also underscore the relationship between mental illness, substance abuse, and
violence. The MacArthur Study followed 951 psychiatric patients for one year
after they were discharged from acute psychiatric units. 220 In contrast to the
Epidemiological Catchment Area Study, researchers used three sources of
information to determine the prevalence of violence—interviews with patients,
interviews with collateral informants (usually a family member), as well as
hospital and arrest records. In the MacArthur study, “violence” included battery
that resulted in a physical injury, sexual assault, assault with a weapon, or
threats with a weapon. Consistent with prior research in this area, substance
abuse emerged as an important risk factor for violence. Among patients with
both an Axis I mental disorder—e.g. schizophrenia, major depression, or
bipolar disorder—and a substance abuse disorder the one-year prevalence of
violence was 31.1%, compared to 17.9% among patients without a substance
abuse disorder.221
Focusing on one study site, Pittsburgh, Pennsylvania, researchers then
compared violence among discharged psychiatric patients to the prevalence of
violence among others living in the same neighborhood. Once again, substance
abuse emerged as a significant risk factor for violence. The study found that
discharged psychiatric patients without a substance abuse problem were no
more likely to engage in violence than other people living in the same
neighborhood without a substance abuse problem. However, the presence of a
substance abuse problem raised the prevalence of violence among discharged
patients and the comparison group. The prevalence of violence among patients
who reported symptoms of substance abuse was significantly higher than the
prevalence of violence among others in their neighborhood.222
Discharged
patients were also more likely to report drug and alcohol abuse than community
controls.223 Contrary to public fears of street violence involving people with
mental illnesses, the targets of violence were usually family members or
SERVICES 654 (2001) (finding that approximately three percent of violent offenses could be attributed to individuals who had a principal diagnosis of any non-substance abuse related mental disorder).
220. Henry J. Steadman et al., Violence by People Discharged from Acute Psychiatric Facilities and by Others in the Same Neighborhoods, 55 ARCHIVES GEN. PSYCHIATRY 393 (1998).
221. Id. at 399.
222. Id. at 400.
223. Id. at 400; see also Eric B. Elbogen & Sally C. Johnson, The Intricate Link Between Violence and Mental Disorder, 66 ARCHIVES GEN. PSYCHIATRY 152 (2009) (finding that incidence of violence was higher for people with severe mental illnesses but only significantly so for people with a co-occurring substance abuse or substance dependence disorder); E. Fuller Torrey, Jonathan Stanley, John Monahan, Henry J. Steadman & the MacArthur Study Group, The MacArthur Violence Risk Assessment Study Revisited: Two Views Ten Years After its Initial Publication, 59 PSYCHIATRIC SERVICES 147 (2008) (debating design of the MacArthur Study). But see Sally Satel & D.J. Jaffe, Violent Fantasies, 50 NAT’L. REV. 36 (1998) (offering a critique of the MacArthur Study).
194 STANFORD LAW & POLICY REVIEW [Vol. 26:159
friends, and most incidents occurred in the home. Nor were there significant
differences between patients and community controls.224
Only a few studies have investigated the number of homicides committed
each year by people with mental illnesses in the United States. Some of them
suggest a similar relationship between substance abuse, mental illness and the
targets of violence. For example, a 1994 study by the Justice Department found
that of the 20,680 homicides reported in 1988, only 4.3% or 889 of those
homicides were committed by someone with a history of mental illness.225 The
same study reported that 64.4% of the defendants charged in those cases were
under the influence of alcohol at the time of the offense. More recent work in
this area suggests that the number of homicides attributable to people with
mental illnesses might be somewhat higher. A 2008 study conducted by
Matejkowski and colleagues found that of the 518 people convicted of murder
in Indiana between 1990 and 2002, 95 of them or 18% had a record diagnosis
of mental illness.226 The most common diagnosis was depression, followed by
schizophrenia and other psychotic disorders. In addition, more than half of all
offenders with a mental disorder also had a history of alcohol and drug abuse.
Most victims were family members or a companion, and most homicides
occurred in the home of the offender or a shared residence of the victim and the
offender.227
b. Psychosis and Violence
Researchers have also investigated the relationship between psychosis and
violence.228 In a 1992 study on the relationship between psychosis and
violence, Bruce Link and colleagues compared arrest rates and self-reported
acts of violence among psychiatric patients, residing in the Washington Heights
area of New York City, to people who had never received mental health
treatment, residing in the same neighborhood.229 The study found that 15% of
the community sample who had never received treatment from a mental health
224. Steadman et al., supra note 220, at 399.
225. JOHN M. DAWSON & PATRICK A. LANGAN, BUREAU OF JUSTICE STATISTICS, U.S. DEP’T OF JUSTICE, MURDER IN FAMILIES 4 (1994).
226. Jason C. Matejowski et al., Characteristics of Persons with Severe Mental Illness Who Have Been Incarcerated for Murder, 36 J. AM. ACAD. PSYCHIATRY & L. 74, 76 (2008).
227. Id. at 79.
228. Psychosis is not itself a mental disorder, but rather a general term for a cluster of psychiatric symptoms, primarily delusions and hallucinations. The Diagnostic and Statistical Manual of Mental Disorders defines delusions as “fixed beliefs that are not amenable to change in light of conflicting evidence.” AM. PSYCHIATRIC ASS’N, DIAGNOSTIC AND
STATISTICAL MANUAL OF MENTAL DISORDERS: FIFTH EDITION 87 (2013). Hallucinations are perceptions without an external stimulus. Id.. Common delusions include the belief that thoughts have been removed from the patient’s mind by an outside force (thought withdrawal) and that thoughts have been placed into the patient’s mind (thought insertion).
229. Bruce G. Link et al., The Violent and Illegal Behavior of Mental Patients Reconsidered, 57 AM. SOC. REV. 275, 278 (1992).
2015] INVOLUNTARY OUTPATIENT COMMITMENT 195
professional reported fighting within the last five years, compared to 29% of
repeat psychiatric patients.230
Link and colleagues then controlled for psychotic
symptoms using a scale derived from the Psychiatric Epidemiology Research
Interview (PERI). After controlling for psychotic symptoms, the study found
that being a repeat psychiatric patient was no longer a statistically significant
predictor of violence. Instead, much of the difference between psychiatric
patients and community controls could be explained by the level of psychotic
symptoms.231 Moreover, even among residents who had never been treated for
a psychiatric disorder, psychotic symptoms were associated with violence.
In further analyses, Link and colleagues found that three symptoms of
psychosis termed “threat/control-override” symptoms were associated with
significant increases in violent behavior, even after controlling for
sociodemographic variables and other psychotic symptoms.232 The symptoms
included feeling that your mind has been dominated by forces beyond your
control, that thoughts put into your head were not your own, and that people
wished to do you harm.233 Similarly, using data from the Epidemiological
Catchment Area Study, Jeffrey Swanson and colleagues found that respondents
who reported one or more threat/control-override symptoms were more than
twice as likely to report violence during the previous year, compared to
respondents who reported other psychotic symptoms, and six times more likely
to report violence compared to people without a mental disorder.234
However, as Link and colleagues caution, when compared to the risk of
violence associated with age, gender and socioeconomic status, the risk of
violence associated with psychotic symptomatology is relatively modest.235 For
example, gender was far more predictive of arrests, fighting, and ever hurting
someone badly than status as a former or repeat psychiatric patient.236
Moreover, not all studies have found an association between psychotic
symptoms and violence. Some studies, including analyses using ECA data,
have not found an association between TCO symptoms and violence when
230. Id. at 283.
231. Id. at 288.
232. Bruce G. Link & Ann Stueve, Psychotic Symptoms and the Violent/Illegal Behavior of Mental Patients Compared to Community Controls, in VIOLENCE AND MENTAL
DISORDER 137 (John Monahan & Henry J. Steadman eds., 1994).
233. Id. at 143; see also Bruce Link & Ann Stueve, Psychotic Symptoms and Violent Behaviors: Probing the Components of the “Threat/Control-Override” Symptoms, 33 SOC. PSYCHIATRY & PSYCHIATRIC EPIDEMIOLOGY S55 (1998) (finding that both threat and control-override delusions predicted were independently associated with violent behavior).
234. Jeffrey W. Swanson et al., Psychotic Symptoms and Disorders and the Risk of Violent Behaviour in the Community, 6 CRIM. BEHAV. & MENTAL HEALTH 309 (1996).
235. Link et al., supra note 229, at 290.
236. See also Eric B. Elbogen & Sally Johnson, The Intricate Link Between Violence and Mental Disorder, 66 ARCHIVES OF GEN. PSYCHIATRY 152 (2009) (finding that mental illness alone did not predict future violence; however, age, gender, having completed less than a high school education, a history of violence, and a history of juvenile detention, accounted for one quarter of the variance in violent behavior).
196 STANFORD LAW & POLICY REVIEW [Vol. 26:159
controlling for the effects of treatment noncompliance or the presence of a
substance abuse disorder.237 Using data from the MacArthur Study, Paul
Appelbaum and colleagues found that delusions were not associated with an
increased risk of violent behavior, nor were particular threat/control-override
(TCO) symptoms associated with a greater risk of violence.238 To the contrary,
the study found that patients with threat/control-override delusions were
significantly less likely to engage in violence than patients without similar
delusions. Appelbaum and colleagues note that people who experience chronic
psychosis also tend to experience social withdrawal and smaller social
networks. The authors hypothesize that with smaller social networks, people
who experience chronic psychosis might have fewer interpersonal interactions,
and thus fewer relationships that lead to violence.239
2. Risk Assessment and Violence
Most studies suggest that by itself mental illness is at best a poor predictor
of violence. Instead, situational and demographic factors such as being young,
male, and unemployed with a history of violence appear to be better
predictors.240 Even then, predictions of violence are notoriously difficult.
237. Jeffrey Swanson et al., Violence and Severe Mental Disorder in Clinical and Community Populations: The Effects of Psychotic Symptoms, Comorbidity, and Lack of Treatment, 60 PSYCHIATRY 1 (1997).
238. Paul Appelbaum et al., Violence and Delusions: Data from the MacArthur Violence Risk Assessment Study, 157 AM. J. OF PSYCHIATRY 566 (2000); see also Thomas Stompe et al., Schizophrenia, Delusional Symptoms, and Violence: The Threat/Control-Override Concept Reexamined, 30 SCHIZOPHRENIA BULL. 31 (2004) (finding that threat/control-override symptoms were not associated with violence in a retrospective study of male offenders with schizophrenia).
239. See Appelbaum et al., supra note 237; see also Louise G. Braham et al., Acting on Command Hallucinations and Dangerous Behavior: A Critique of the Major Findings in the Last Decade, 24 CLINICAL PSYCHOL. REV. 513 (2003) (noting the relationship between command hallucinations and violence); Sue Estroff et al., The Influence of Social Networks and Social Support on Violence by Persons with Serious Mental Illness, 45 HOSP. &
COMMUNITY PSYCHIATRY 669 (finding that respondents with larger networks, and those whose networks primarily consisted of relatives, were more likely to threaten violence). One might think when people with mental illnesses, like schizophrenia, hear command hallucinations they comply unthinkingly. However, the relationship between command hallucinations and violence is a complicated one, mediated by multiple psychological processes, including beliefs about the voice, the content of the command, and the consequences of noncompliance.
240. See Jeffrey Swanson et al., Violent Behavior Preceding Hospitalization Among Persons with Severe Mental Illness, 23 LAW & HUM. BEHAV. 185 (noting that, while clinical diagnosis and symptom variables were not significantly associated with violence, violent behavior among revolving door patients was associated with substance abuse, young age, a history of victimization); Jeffrey Swanson et al., Violence and Psychiatric Disorder in the Community: Evidence from the Epidemiologic Catchment Area Surveys, 41 HOSP. &
COMMUNITY PSYCHIATRY 761, 763 (1990) (finding that being young, male, and of a low socioeconomic status was associated with violent behavior); see also Jeffrey Swanson et al., The Social-Environmental Context of Violent Behavior in Persons Treated for Severe Mental Illness, 92 AM. J. PUB. HEALTH 1523 (2002) (noting that “[p]sychopathology per se seldom
2015] INVOLUNTARY OUTPATIENT COMMITMENT 197
Owing to the low base rate of violent crime, even the best methods will
produce a large number of false positives. For example, a well-known study on
risk assessment and violence, conducted by Charles Lidz and colleagues, found
that clinical predictions of violence were only slightly better than chance.241
The study involved two groups of psychiatric patients matched on age, race,
gender, and admission status. One group consisted of patients who were
predicted to be violent by clinicians; the other was a comparison group of
patients who were not predicted to be violent.
Lidz and colleagues assessed the accuracy of clinical judgments by
calculating the sensitivity and specificity of their predictions. Sensitivity (the
true positive rate) measures the proportion of true positives that are correctly
identified. Specificity (the true negative rate) measures the proportion of true
negatives that are correctly identified. Clinicians accurately identified 60% of
patients who turned out to be violent and 58% of patients who turned out to be
nonviolent.242 As a result, the study reported a considerable number of false
negatives and false positives. 129 patients who were not predicted to be violent
were in fact violent, and 167 patients who were predicted to be violent during
the study period did not engage in violence.243
A second approach to risk assessment uses statistical or actuarial methods
to assess the risk of violence. Clinical approaches to risk assessment depend on
a clinician to estimate the risk of violence based on his or her clinical judgment;
however, actuarial methods are based on a standardized list of validated risk
factors, such as age, gender, and past history of violence. Most studies have
shown that actuarial methods tend to be more accurate than clinical
predictions.244 For example, using data from the Lidz study, William Gardner
and colleagues found that actuarial methods had lower false positive rates and
lower false negative rates when compared to clinical prediction.245 Actuarial
predictions based only on the patient’s history of violence were also more
leads to assaultiveness,” but may converge with other risk factors such as violent victimization and exposure to violence to increase the risk of violent behavior); Estroff et al., supra note 238; Virginia Hiday, The Social Context of Mental Illness and Violence, 36 J. HEALTH & SOC. BEHAV. 122 (1995).
241. Charles W. Lidz et al., The Accuracy of Predictions of Violence to Others, 269 J. AM. MED. ASS’N 1007 (1993).
242. Id. at 1009. Sensitivity or the true positive rate of 60% equals the number of true positives (i.e., the 190 patients who were predicted to be violence and were in fact violent) divided by the sum of the true positives and the false negatives (i.e., the 129 patients who were not predicted to be violent but were violent). Specificity or the true negative rate of 58% equals the number of true negatives (i.e., 228 comparison patients who were not violent) divided by the sum of the true negatives and the false positives (the 167 patients who were predicted to be violent but were not violent).
243. Id.
244. See William M. Grove & Paul Meehl, Comparative Efficiency of Informal (Subjective, Impressionistic) and Formal (Mechanical, Algorithmic) Prediction Procedures: The Clinical-Statistical Controversy, 2 PSYCHOL. PUB. POL’Y. & L. 293 (1996).
245. See William Gardner et al., Clinical Versus Actuarial Predictions of Violence in Patients with Mental Illnesses, 64 J. CONSULTING & CLINICAL PSYCHOL. 602 (1996).
198 STANFORD LAW & POLICY REVIEW [Vol. 26:159
accurate than clinical predictions of violence. Using only the patient’s history
of violence, an actuarial model was able to identify 71% of patients who were
violent, while clinical methods only identified 62% of patients who engaged in
violence.246
Although several studies have shown that actuarial risk assessments tend to
be more accurate than clinical predictions, most outpatient commitment laws
do not require actuarial assessments. Nor are actuarial methods often used,
since they can be time-consuming and cumbersome.247 Moreover, as David
Cooke and Christine Michie have observed, “it is a statistical truism that “the
mean of a distribution tells us about everyone, [and] yet no one.”248 Actuarial
assessments estimate the likelihood of future violence based on the behavior of
a group. However, any significant social, psychological, or environmental
differences between the individual and the group can increase or decrease the
likelihood of violence.249 Further, studies consistently show that a history of
violence, and, in particular, a recent overt act of violence, are among the best
predictors of future violence.250 However, most outpatient commitment laws do
not require a recent act of violence. For example, under Kendra’s Law, a court
may order outpatient commitment if, in addition to proof on all other elements,
a history of treatment noncompliance has resulted in one of more acts of
violence toward others within the last forty-eight months, or even a threat or
attempt at serious physical harm toward others within the last forty-eight
months.251
For people with serious and persistent mental illnesses, being misclassified
as dangerous can have serious consequences. Even when participation in an
assisted outpatient treatment program offers a less restrictive alternative to
hospitalization, a court order to participate in group therapy burdens the liberty
interests of persons who are predicted to be violent, but who are in fact not
violent. Courts will sometimes order an assertive community treatment team to
visit the patient’s home, further burdening the outpatient’s interest in privacy.
246. Id. at 607.
247. See John Monahan et al., Developing a Clinically Useful Actuarial Tool for Assessing Violence Risk, 176 BRIT. J. PSYCHIATRY 312 (2000).
248. David J. Cooke & Christine Michie, Limitations of Diagnostic Precision and Predictive Utility in the Individual Case: A Challenge for Forensic Practice, 34 LAW & HUM
BEHAVIOR 259, 259 (2010); see Jennifer L. Skeem & John Monahan, Current Directions in Violence Risk Assessment, 20 PSYCHOL. MED. 38 (2011), for an argument that group data can be useful for individual cases; see also David J. Cooke & Christine Michie, Violence and Risk Assessment: Challenging the Illusion of Certainty, in DANGEROUS PEOPLE: POLICY, PREDICTION, AND PRACTICE 147 (Bernadette McSherry & Patrick Keyzer eds., 2011).
249. See JOHN PARRY, CIVIL MENTAL DISABILITY LAW, EVIDENCE AND TESTIMONY 669 (2010).
250. See Douglass Mossman et al., Risky Business Versus Overt Acts: What Relevance Do “Actuarial,” Probabilistic Risk Assessments Have for Judicial Decisions on Involuntary Psychiatric Hospitalizations?, 11 HOUS. J. HEALTH L. & POL’Y 365 (2011).
251. N.Y. MENTAL HYG. LAW § 9.60(c)(2)(4)(i)-(ii) (McKinney, Westlaw through 2014 Sess.).
2015] INVOLUNTARY OUTPATIENT COMMITMENT 199
There are also system wide costs. With an overemphasis on dangerousness,
states risk diverting limited resources toward programs for people with mental
illnesses who are thought to be dangerous, and away from the majority of
people with mental illnesses who are not dangerous.
3. The Criminal-Civil Distinction
Suppose, however, that at least when combined with a substance abuse
disorder, mental illness gives us good reason to suspect a heightened risk of
violence. In addition, suppose we are able to predict violence to a reasonable
degree of certainty. What justifies assisted outpatient treatment based on our
suspicion that, at some point in the future, the subject of a court order might
harm others before he has actually done so? As a general matter the other-
regarding harms we aim to prevent through outpatient commitment are
addressed through the deterrent and retributive functions of the criminal justice
system.
The Supreme Court has yet to address the constitutionality of outpatient
commitment. However, in Kansas v. Hendricks and Kansas v. Crane the Court
considered an analogous problem that arises when states use civil commitment
to detain sex offenders beyond the expiration of their criminal sentences.252 In
both cases, the Court narrowed the class of offenders eligible for civil
commitment to those whose “mental abnormality” rendered them dangerous
beyond their control. Writing for the Court in Crane, Justice Breyer averred:
It is enough to say that there must be proof of serious difficulty in controlling
behavior. And this, when viewed in light of such features of the case as the
nature of the psychiatric diagnosis and the severity of the mental abnormality
itself, must be sufficient to distinguish the dangerous sexual offender whose
serious mental illness, abnormality, or disorder subjects him to civil
commitment from the dangerous but typical recidivist convicted in an ordinary
criminal case.253
Echoing Hendricks, Justice Breyer added that the distinction is a necessary
one “lest civil commitment [should] become a mechanism for retribution or
general deterrence—functions properly those of criminal law, not civil
commitment.”254 In both Hendricks and Crane, the Supreme Court reaffirmed
the criminal justice system as the preferred approach to garden variety criminal
conduct. The underlying assumption of the criminal law is that most of us have
at least a normal capacity to understand what the law requires, and most of us
have at least a normal capacity to order our conduct within the wide boundaries
set by legal norms. When culpable agents breach legal norms of their own
volition, we describe their conduct as a “crime” meriting “punishment” rather
252. Kansas v. Hendricks, 521 U.S. 346 (1997); Kansas v. Crane, 534 U.S. 407 (2002).
253. Crane, 534 U.S. at 413.
254. Id. at 412 (internal quotations omitted).
200 STANFORD LAW & POLICY REVIEW [Vol. 26:159
than a “breach” creating “liability.”255 In doing so, we communicate
reprobation for wrongdoing while also addressing the offender as a moral
agent. By contrast, the moral legitimacy of civil commitment rests on its
limitation to persons who lack the capacities for moral responsibility or
criminal responsibility. As Allen Buchanan and Dan Brock write, “[i]f the
dangerous mentally ill are justifiably treated differently, it must be because they
are not capable of responsibly controlling their behavior that is dangerous to
others as required by criminal prohibitions.”256
In the same way, the Supreme Court limited sex offender commitments to
those whose mental abnormalities rendered them unable to control their
behavior. Even so, critics argue that the Court’s inability-to-control formulation
is vastly overbroad and unworkable. As Christopher Slobogin writes, “evidence
that the impulses experienced by addicts, sexual offenders and people with
psychosis are stronger than those that lead people to commit typical crime is
hard to come by; burglars recidivate at least as much as sex offenders, and
white collar criminals are probably just as ‘driven’ by urges, albeit for things
like wealth, fame or power rather than (or perhaps in addition to) drugs or
sex.”257 To that end, a second approach rejects the volitional impairment
approach entirely. Adherents to this view—foremost among them, Eric Janus,
Robert Schopp and Stephen Morse—argue that police power commitments are
appropriate, but only for those who are, in essence, “too sick to deserve
punishment.”258 As Stephen Morse writes, “[f]or reasons much studied and
theorized about, but in fact not very well understood, some unfortunate people
are so irrational, so grossly out of touch with reality, that ascribing
responsibility to them is a travesty according to any but the most extravagantly
libertarian account of human agency.”259 If under the grip of delusional beliefs
such an agent were to strike out at a perceived threatener, she would not be
morally responsible for her actions and therefore not deserving of legal
punishment.
My own view, to be developed in Part IV, rests on a combination of both
approaches. Under certain circumstances, outpatient commitment may be
appropriate for people with mental illnesses who are irrational in the way
255. Paul H. Robinson, Foreword: The Criminal-Civil Distinction and Dangerous Blameless Offenders, 83 J. CRIM. L. & CRIMINOLOGY, 693 (1993).
256. ALLEN E. BUCHANAN & DAN W. BROCK, DECIDING FOR OTHERS: THE ETHICS OF
SURROGATE DECISION MAKING 329 (1990).
257. CHRISTOPHER SLOBOGIN, MINDING JUSTICE: LAWS THAT DEPRIVE PEOPLE WITH
MENTAL ILLNESSES OF LIFE AND LIBERTY, 131 (2006).
258. Eric S. Janus, Hendricks and the Moral Terrain of Police Power Civil Commitment, 4 PSYCHOL. PUB. POL’Y & L. 297, 298 (quoting Millard v. Harris, 406 F.2d 964, 969 (D.C. Cir. 1968)); accord ROBERT F. SCHOPP, COMPETENCE, CONDEMNATION AND
COMMITMENT 148-51 (2001); Robert F. Schopp, Outpatient Civil Commitment: A Dangerous Charade or a Component of a Comprehensive Institution of Civil Commitment?, 9 PSYCHOL. PUB. POL’Y & L. 33, 46 (2003).
259. Stephen Morse, Blame and Danger: An Essay on Preventive Detention, 76 B.U. L. REV. 113, 123 (1996).
2015] INVOLUNTARY OUTPATIENT COMMITMENT 201
Morse suggests. Alternatively, outpatient commitment might be appropriate for
people with mental illnesses who are unable to control their behavior. Before
offering an alternative approach to outpatient commitment determinations, Part
III will press on to examine further justifications for outpatient commitment.
B. Harm to Self
Violent crimes involving people with mental illnesses have fueled an
interest in preventive outpatient commitment. Yet supporters of outpatient
commitment have also argued that court orders to participate in treatment are
amply justified by the risk of serious harms to oneself.260 On any given night in
the United States, roughly 650,000 people are homeless, and 1.5 million are
homeless at some point during the year.261 An estimated twenty-six percent of
sheltered persons who are homeless have a severe mental illness.262 For people
with severe and persistent mental illnesses, the failure to comply with
prescribed medications can increase the risk of homelessness.263 Although we
tend to think of people with mental illnesses as much more likely to commit
violent crimes than others, studies have shown that people with mental illnesses
are actually far more likely to be the victims of violent crime, rather than the
perpetrators of violent crime, when compared to members of the general
public.264 It may be that psychiatric symptoms such as thought disorganization,
impaired reality testing, poor impulse control, and poor problem solving
abilities impeded the ability to perceive risks and protect oneself.265
For others, treatment noncompliance will lead predictably to incarceration,
often for nonviolent offenses such as trespass, loitering, and disorderly
conduct.266 Of the nearly 2 million inmates held in jails and prisons, an
260. John Kip Cornwell & Raymond Deeney, Exposing the Myths Surrounding Outpatient Commitment for Individuals with Chronic Mental Illness, 9 PSYCHOL. PUB. POL’Y
& L. 209 (2003); Kress, supra note 2.
261. U.S. DEP’T OF HOUS. & URBAN DEV., THE 2010 ANNUAL HOMELESS ASSESSMENT
REPORT TO CONGRESS i-ii (2010).
262. Id. at 18. By comparison, the National Institute of Mental Health estimates that 4.1% of all adults in the United States aged 18 or older—approximately 9.6 million adults—had a serious mental illness in 2012. See Serious Mental Illness (SMI) Among U.S. Adults, NIMH, http://www.nimh.nih.gov/health/statistics/prevalence/serious-mental-illness-smi-among-us-adults.shtml (last visited Nov. 4, 2014).
263. Mark Olfson, Predicting Medication Noncompliance After Hospital Discharge Among Patients with Schizophrenia, 51 PSYCHIATRIC SERVICES 216, 220 (2000).
264. Linda A. Teplin et al., Crime Victimization in Adults with Severe Mental Illness: Comparison with the National Crime Victimization Survey, 62 ARCHIVES GEN. PSYCHIATRY 911 (2005) (finding that more than one quarter of persons with a serious mental illness had been victims of a violent crime in the past year, a rate more than eleven times higher than the general population rates even after controlling for demographic factors).
265. Id.
266. NAT’L ASS’N OF MENTAL HEALTH PLANNING & ADVISORY COUNCILS, SUBSTANCE
ABUSE & MENTAL HEALTH SERVS. ADMIN., JAIL DIVERSION: STRATEGIES FOR PERSONS WITH
202 STANFORD LAW & POLICY REVIEW [Vol. 26:159
estimated 300,000 suffer from a major mental illness.267 Left untreated, people
with severe and persistent mental illnesses are also more likely to commit
suicide.268
The legislative findings for Kendra’s Law suggest that an important
purpose of the assisted outpatient treatment program is to improve the
wellbeing of people with mental disorders:
The legislature . . . finds that some mentally ill persons, because of their
illness, have great difficulty taking responsibility for their own care, and often
reject the outpatient treatment offered to them on a voluntary basis . . . . The
legislature therefore finds that assisted outpatient treatment as provided in this
act is compassionate, not punitive, will restore patients’ dignity, and will
enable mentally ill persons to lead more productive and satisfying lives.269
Supporters of involuntary outpatient commitment view these laws as a
compassionate, commonsense response to the symptoms of a failing mental
health system—homelessness, victimization, incarceration, and suicide.270
Opponents view them as “a glaring example of paternalism gone awry.”271 The
difference between them goes to a longstanding conflict between champions of
liberalism and their critics. For strong supporters of involuntary outpatient
commitment—like critics of civil commitment reform before them—the idea
that people with mental illnesses should be “free to rot” seems unfathomable.
Yet liberal commitments to neutrality limit the power of governments to
enforce any particular conception of the good life. Parallel commitments to
autonomy and personal sovereignty limit the power of governments to prevent
citizens from harming no one other than themselves. The same liberal
commitments to autonomy led the New York Court of Appeals to uphold the
right of a competent person to refuse treatment in Rivers v. Katz,
notwithstanding a diagnosis of mental illness.
The question for liberalism is this: when—if ever—is paternalism
justified? I shall define paternalism in Gerald Dworkin’s terms. On his view, A
behaves paternalistically toward B by doing (or omitting) C if—(i) C (or its
omission) interferes with the liberty or autonomy of B; (ii) A does so without
SERIOUS MENTAL ILLNESS 4 (2005), available at http://www.namhpac.org/ PDFs/01/jaildiversion.pdf.
267. Fred E. Markowitz, Mental Illness, Crime and Violence: Risk, Context and Social Control, 16 AGGRESSION & VIOLENT BEHAV. 36, 38 (2011); see also PAULA M. DITTION, BUREAU OF JUSTICE STATISTICS, SPECIAL REPORT: MENTAL HEALTH AND TREATMENT OF
INMATES AND PROBATIONERS (1999).
268. Brian Kenneth Ahmedani et al., Suicide Thoughts and Attempts and Psychiatric Treatment Utilization: Informing Prevention Strategies, 63 PSYCHIATRIC SERVICES 186 (2012).
269. Mental Hygiene—Community-Based Settings—Kendra’s Law, ch. 408, S.5762-A, 1999 N.Y. Sess. Law (codified as amended at N.Y. MENTAL HYG. LAW § 9.60 (McKinney, Westlaw through 2014 Sess.)).
270. See E. FULLER TORREY, THE INSANITY OFFENSE: HOW AMERICA’S FAILURE TO
TREAT THE SERIOUSLY MENTALLY ILL ENDANGERS ITS CITIZENS (2008); Torrey & Zdanowicz, supra note 14, at 340.
271. Munetz, et al., supra note 16, at 145.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 203
the consent of B; and (iii) A does so because A believes C will improve the
welfare or in some way promote the interests, values, or good of B.272 In short,
for Dworkin, paternalism involves interfering with another person, against her
will, because doing so will make her better off. We can distinguish two
varieties of paternalism—hard and soft. Hard paternalism is the view that
governments may intervene to protect competent adults from the harmful
consequences of their fully voluntary self-regarding behavior.273 Soft
paternalism holds that states may intervene to prevent harmful self-regarding
conduct when, and only when, that conduct is substantially nonvoluntary, or
when intervention is necessary to determine whether the conduct is
substantially nonvoluntary.274 Non-state actors may behave paternalistically, to
be sure, but my concern here is with the moral limits on government action.
As philosopher Joel Feinberg observed, “[p]aternalism is something we
often accuse people of,” in large part because paternalism is usually thought to
show inadequate respect for personal autonomy.275 Yet, given the context of
our discussion, two questions immediately arise. What do we mean by
autonomy? Further, is autonomy something we have reason to value when a
person has a serious mental illness? To oppose involuntary mental health
treatment on civil liberty or autonomy grounds is, as one commentator writes, a
case of “autonomy gone mad.”276 To be autonomous in any real sense—in any
sense that we are obligated to respect—presupposes a capacity for rational
understanding or rational choice, capacities that are significantly diminished
when a person is severely and persistently mentally ill. Others question whether
there are real liberty interests to contend with here at all:
We argue that the real liberty question regarding individuals with severe
psychiatric disorders is whether they are in fact free when ill. If one’s thoughts
and behavior are driven by a disease process of the brain over which one has
no control, is this truly liberty? . . . . “Medication can free victims from their
illnesses—free them from the Bastille of their psychoses—and restore their
dignity, their free will, and the meaningful exercise of their liberties.” 277
Part III.B examines these questions. I begin by expounding Feinberg’s
conception of autonomy. I then defend his approach against a few criticisms. I
argue that Joel Feinberg’s theory of autonomy has sufficient room to
accommodate people with mental illnesses. Nor is it incoherent to claim that
persons with mental illnesses have interests in autonomy that we are obligated
to respect. I then turn to the justifications for paternalistic intervention. My aim
here will be to clarify the meaning of personal autonomy and the moral basis
272. Gerald Dworkin, Paternalism, STAN. ENCYCLOPEDIA PHIL., http://plato. stanford.edu/entries/paternalism (offering Dworkin’s most recent definition of paternalism).
273. 3 JOEL FEINBERG, THE MORAL LIMITS OF THE CRIMINAL LAW 12 (1986)
274. Id.
275. Id. at 4.
276. Alfred Tauber, Autonomy Gone Mad, 10 PHIL. CONTEMP. WORLD 75 (2003).
277. Torrey & Zdanowicz, supra note 14, at 341 (quoting Herschel Hardin, Uncivil Liberties, VANCOUVER SUN, July 22, 1993, at A15).
204 STANFORD LAW & POLICY REVIEW [Vol. 26:159
for paternalistic intervention and, in doing so, clarify our thinking for the hard
cases that will inevitably arise when a person refuses mental health treatment.
1. Autonomy Defined
In his seminal treatise on paternalism, Harm to Self: The Moral Limits of
the Criminal Law, Joel Feinberg elucidates four conceptions of autonomy:
autonomy as the capacity to govern oneself; autonomy as the actual condition
of self-government; autonomy as a character ideal; and autonomy as the
sovereign authority to govern oneself.278
Feinberg understands autonomy to involve both the capacity for self-
government and the actual condition of self-government. He defines the
capacity for self-government as the capacity to make rational choices.279 For
Feinberg and others, the capacity to make rational choices is a threshold
concept. “Some competent persons are no doubt more richly endowed with
intelligence, judgment, and other relevant capabilities than others,” he writes,
“but above the appropriate threshold they are deemed no more competent
(qualified) than the others at the ‘task’ of living their own lives according to
their own values as they choose.”280
In Feinberg’s view, a person might possess the capacity for self-
government—insofar as he has the capacity to make rational choices—but be
less than fully autonomous because he does not actually govern himself. There
are many ways in which a person might fall short of the actual condition of
self-government. “I do not govern myself if you overpower me by brute force,”
or if “poverty reduces me to abject dependence on the assistance of others.”281
Nor am I autonomous if I am no more than “the mouthpiece of other persons or
forces.”282 He writes, “[a]utonomy, so understood, refers to a congeries of
virtues,” among them, a person’s interest in being “his own man,” or “her own
woman,” her interest in maintaining a distinct self-identity, and importantly,
personal authenticity.283
A person is authentic to the extent that, unlike both the inner-directed and the
other-directed person, he can and does subject his opinions and tastes to
rational scrutiny. He is authentic to the extent that he can and does alter his
convictions for reasons of his own, and does this without guilt or anxiety.284
A conception of autonomy, Feinberg adds, can include the notion of
autonomy as “an attractive ideal for human character,”285 with the result that “it
278. FEINBERG, supra note 273, at 27-51.
279. Id. at 28.
280. Id. at 30.
281. Id. at 31.
282. Id. at 32.
283. Id. at 31-32.
284. Id. at 33.
285. Id. at 31.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 205
is better to be autonomous than not,” all other things being equal.286 In the
same way, others have said that a person might value autonomy because she
derives pleasure from exercising her capacity for autonomy.287 Or a person
might value autonomy because she derives self-respect from being recognized
by others as “the kind of creature” who is capable of exercising autonomy.288
Feinberg understands the core of autonomy as the sovereign authority to
govern oneself.
The life that a person threatens by his own rashness is after all his life; it
belongs to him and to no one else. For that reason alone, he must be the one to
decide—for better or worse—what is to be done with it in that private realm
where the interests of others are not directly involved.289
So construed, autonomy involves the sovereign right of self-determination
within the boundaries of the personal domain, not unlike the sovereign rights of
an autonomous nation-state within its own borders. “My personal domain,”
Feinberg writes, consists of the body, privacy interests and “critical life-
decisions,” such as what to study or what to career to pursue, and all decisions
that are chiefly or primarily self-regarding.290
From John Stuart Mill, Feinberg borrows a useful, albeit imperfect,
distinction between self-regarding and other-regarding harm. With minor
alternations, I shall rely these definitions as well. Self-regarding harms are
those which “chiefly,” “primarily,” or “directly” frustrate the interests of the
decision-maker.291 Beyond the personal domain are other-regarding harms
which “directly and in the first instance affect the choices and sensibilities of
other persons.”292
2. Conceptual Hurdles and Replies
Feinberg’s account privileges authenticity as an element of autonomy. “To
the degree to which a person is autonomous he is not merely the mouthpiece of
other persons or forces,” he observes.293 “Rather his tastes, opinions, ideals,
goals, values and preferences are all authentically his.”294
286. Id. at 45. Here Feinberg anticipates the critique from communitarianism: “The ideal of the autonomous person is that of an authentic individual whose self-determination is as complete as is consistent with the requirement that he is, of course, a member of a community.” Id. at 47.
287. See GERALD DWORKIN, THE THEORY AND PRACTICE OF AUTONOMY 112 (1988) (“[T]he process of thinking about, reflecting upon, choosing among preferences is a source of satisfaction to individuals.”).
288. Id.
289. FEINBERG, supra note 273, at 59.
290. Id. at 54-55.
291. Id. at 56.
292. Id.
293. Id. at 32.
294. Id.
206 STANFORD LAW & POLICY REVIEW [Vol. 26:159
Feinberg’s conception of autonomy as authenticity has natural purchase in
a discussion of mental illness. Consider first a person who is intoxicated. After
party, it might be quite natural to say, “Please excuse my friend. He has had too
much to drink this evening and he is not himself.” The claim here is that the
person’s behavior does not reflect his true self, but rather the influence of
alcohol.295 In the same way, people with mental illnesses sometimes
experience their symptoms as alien or ego dystonic. In the years before he
pushed Kendra Webdale to her death, Andrew Goldstein believed that he was
being controlled by someone (named Larry) who inhabited his body and
removed his brain.296 When a person refuses treatment in the grip of such a
delusion, we will wonder whether his treatment preferences are truly his own.
Authenticity conditions are common in the literature on autonomy.297 In Part
IV.A, I will argue that the authenticity problem explains much of our dilemma
when a person refuses treatment for unusual reasons.298
Still, an account of autonomy that incorporates a notion of authenticity
must overcome a few challenges. First, in what sense does autonomy demand
that we alter our convictions for reasons of our own? If we take authenticity to
require that all of our beliefs are entirely our own, then as Feinberg concedes,
“nothing resembling rational reflection can ever get started.”299 The
requirement that we engage in rational reflection on our convictions is too
demanding. Nonetheless, Feinberg has a more modest claim in mind—through
a process of critical reflection, the reasons for our actions can become our own.
In this way authenticity arises through a process of self-creation and “self-re-
creation.”300
There is a further sense in which we might question whether it is
appropriate to describe tastes, preferences, goals, or values as authentically
one’s one. Although at times I may act owing to greed, vanity, or foolishness, I
do not endorse these character traits. Indeed I may see them as unfortunate
character flaws that I would rather not have. Yet, when greed, vanity, and
foolishness supply reasons for my actions, are those actions no longer
autonomous, insofar as I am inclined to deny that these reasons are
authentically my own? Here we can distinguish theories in which autonomy
requires second-order identification with first order desires, from accounts in
which authenticity requires merely that the agent would acknowledge (or at
295. NOMY ARPALY, UNPRINCIPLED VIRTUE: AN INQUIRY INTO MORAL AGENCY 123 (2003).
296. CHARLES PATRICK EWING, INSANITY: MURDER, MADNESS AND THE LAW 115 (2008).
297. See Natalie Stoljar, Theories of Autonomy, in PRINCIPLES OF HEALTH CARE ETHICS 11, 14 (R.E. Aschroft et al. eds., 2007); Timothy Kolke, Procedural vs. Substantive Theories of Autonomy: Reinterpreting the Connection Between Good Values and Autonomy, PROLEGOMENA 1 (2003).
298. See infra notes 383-400 and accompanying text.
299. FEINBERG, supra note 273, at 33.
300. Id. at 35 (emphasis in original).
2015] INVOLUNTARY OUTPATIENT COMMITMENT 207
least others would acknowledge) that the reasons for the agent’s actions were
not at all alien to him, but “for better or worse part of the way he is.”301 For
Feinberg, authenticity involves the latter. The result is that for Feinberg, the
excuse we make for our drunken friend is merely a metaphor, while delusions
reported by Andrew Goldstein would be sufficient to cast doubt on his
treatment preferences.
Others might respond that autonomy does not require authenticity at all.
For Gerald Dworkin, what matters is not that my reasons are my own, but
rather than I have the capacity to question whether I will identify with or reject
the reasons for my actions.302 The relevant capacity is the ability to engage in
second-order reflection on our first order motivations and the ability to choose
those motivations if desired.303 By “identify with,” I understand Dworkin to
mean that a person identifies with the reasons underlying her actions when she
has made those reasons her own, or comes to see them as her own.304 In
Dworkin’s view, a conception of autonomy worthy of our respect must meet
several other constraints, foremost among them, a requirement of procedural
independence.305 A person would not count as autonomous under Dworkin’s
view if he has been hypnotized into identifying with his first-order motivations,
otherwise coerced, manipulated, or subliminally influenced. While Dworkin
rejects authenticity or any other substantive rider on autonomy, by itself, a
procedural account of autonomy is not one we ought to adopt.
For Dworkin, a conception of autonomy that insists on a condition of
substantive independence such as authenticity would be inconsistent with other
values we hold: loyalty, objectivity, commitment, benevolence, and love.306
Dworkin thinks that if a person wants to conduct his life in accordance with the
principle that he should do what others tell him to do, we should not, by virtue
of that reason, fail to count him as autonomous. To use Dworkin’s example, if a
person wants to conduct his life in accordance with the following—“[d]o
whatever my mother or my buddies or my leader or my priest tells me to do”—
then so long as that person has not been manipulated in ways that violate
procedural independence, we should count that person as autonomous.307
Yet, we would not be hard pressed to reconcile authenticity with a
commitment to each of these values. As Dworkin rightly points out,
301. Compare FEINBERG, supra note 273, at 165, with Harry G. Frankfurt, Freedom of the Will and the Concept of a Person, in FREE WILL 322, 327-30 (Gary Watson ed., 2003) (asserting that autonomy requires second-order identification with first-order desires).
302. DWORKIN, supra note 287, at 15-20.
303. Id. at 20.
304. As in the following: “A person may identify with the influences that motivate him, assimilate them to himself, view himself as the kind of person who wishes to be moved in particular ways. Or, he may resent being motivated in certain ways, be alienated from those influences, prefer to be the kind of person who is motivated in different ways.” Id. at 15.
305. Id. at 16.
306. Id. at 21.
307. Id.
208 STANFORD LAW & POLICY REVIEW [Vol. 26:159
commitments to each of these values—loyalty, objectivity, benevolence and
love—will sometimes require that we put the interests of others ahead of our
own. However, authenticity only requires that our reasons for acting out of
loyalty or benevolence are in some sense our own. In other words, authenticity
only requires that we do not put the interests of others ahead of our own
because our mothers said so or because altruism is in vogue, but rather because,
at some point, upon reflection we realize that we have reason to value loyalty,
objectivity, benevolence, and love, independent of the fact that someone else
told us to do so. The fact that the people we respect—our mothers, our buddies,
our leaders or our priests—steered us in this direction is not inconsistent with
authenticity, or autonomy, nor is it enough.
3. Autonomy and Mental Illness
In Rivers v. Katz, the New York Court of Appeals upheld the right to refuse
treatment on autonomy grounds.308 With Feinberg’s taxonomy in hand, we are
now in a position to understand the meaning (or meanings) of autonomy and
the myriad autonomy interests of people who are subject to outpatient
commitment orders.
Consider the interest in autonomy as sovereignty. Because outpatient
commitment orders are less restrictive than inpatient commitment, and since
these laws do not permit the use of force to administer medication over a
person’s objection, there is a tendency to view the intrusion on personal
sovereignty as “minimal,” and therefore without serious moral consequence.309
Yet sovereignty, as Feinberg declares, is “an all or nothing concept . . . .” In the
political model, a nation’s sovereignty is equally infringed by a single foreign
fishing boat in its territorial waters as by a squadron of fighter jets flying over
its capital city.”310 A person is entitled to sovereign authority over her domain,
no matter how trivial the intrusion may seem to others. We would not describe
the presence of the fishing boat as an inconsequential forfeiture of sovereignty,
any more than we should characterize an order to participate in group
therapy—for several hours a day, several days a week—as a trivial intrusion on
personal autonomy.
For good reason, some who are ordered to participate in outpatient
treatment on wholly paternalistic grounds will feel not merely irked, annoyed,
or inconvenienced, but rather belittled and demeaned. The essence of their
grievance includes both an interest in autonomy as personal sovereignty, as
well as an interest in autonomy as an ideal of character or human flourishing.
308. 67 N.Y.2d 485, 497-98 (1986); see also supra text accompanying note 194.
309. But see Jessica Wilen Berg & Richard Bonnie, When Push Comes to Shove, in COERCION AND AGGRESSIVE COMMUNITY TREATMENT 169, 172-77 (Deborah L. Dennis & John Monahan eds., 1996) (claiming that while outpatient commitment might impose fewer restrictions on physical liberty, “in many ways it may be more intrusive with regard to a person’s privacy interests” owing primarily to warrantless removal procedures).
310. FEINBERG, supra note 273, at 55.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 209
Earlier I said that a person might value autonomy because she wants to be
viewed by others as “the kind of creature” who is capable of exercising
autonomy.311 A respondent who appeared in court to contest a petition to renew
his assisted outpatient treatment order put the point well:
Q: Why do you want to graduate, what is the big deal about graduating?
A: The big deal is I have to successfully accomplish things I need to
accomplish in my life . . . . I want to move forward in my life . . . . I want to
have a plaque that I could show myself I did this.312
On a Wednesday morning in Brooklyn, the respondent appeared in court to
contest the petition, dressed in a jacket and tie, with his partner and their two-
year-old daughter in tow.313 He explained that he would continue taking his
medications and talking with his counselor, and that he would continue group
therapy but no longer wanted to be on a court order:
Q: What’s the difference between being on a Court Order as opposed to not
being on one?
A: I feel like I am in jail . . . . I don’t know the doctors . . . . I just want to . . .
handle what I have to handle and [without] anything above my head. I just
want to handle without a Court Order.314
For the respondent, it mattered immensely that he was able to show himself
and his family that he had “accomplished” something by graduating and
handling his mental health treatment without being perceived as someone who
requires the oversight of a court order. For him, the injury of outpatient
commitment was an injury to his interest in autonomy as an ideal of human
character or excellence, as well as an injury to his interest in personal
sovereignty.
We can expect cases where mental disorder compromises personal
autonomy and casts doubt on the authenticity of a person’s preferences.
Schizophrenia, bipolar disorder, and depression can impair the ability to reason
and, in turn, the capacity for self-government. In the grip of mania, a person
with bipolar disorder might refuse treatment because he has come to see
himself as invincible, with no need for treatment. A person with clinical
depression might refuse treatment because, owing to depression, he sees
himself as beyond hope and his life as no longer valuable.
However, not all of the reasons a person might offer for refusing treatment
are so obviously distorted by mental illness. As Elyn Saks writes, when a
person refuses treatment, it may be that she is expressing “a legitimate
preference for the symptoms over the cure.”315 Anti-psychotic medications are
311. DWORKIN, supra note 287, at 112.
312. In the Matter of the Application of Adam Karpati, Director of Community Services of the Dep’t of Health and Mental Hygiene for an Order Authorizing and Additional Period of Assisted Outpatient Treatment for John Doe at 29 (N.Y. Sup. Ct. 2010) (No. 300202/10) (transcript on file with author).
313. The author was in the courtroom during the hearing.
314. In the Matter of the Application of Adam Karpati, supra note 312, at 31.
315. Elyn R. Saks, Competency to Refuse Treatment, 69 N.C. L. REV. 945, 990 (1991).
210 STANFORD LAW & POLICY REVIEW [Vol. 26:159
effective, but many of them have adverse side effects including sedation, sexual
dysfunction, and weight gain, as well as an elevated risk of cardiovascular
disease and diabetes.316 People who are able to function in spite of their illness
might oppose additional orders of assisted outpatient treatment because they
find it difficult to go to work or school when they are required to participate in
a partial day program for several days a week. Some might not like their
clinicians, while others who have already participated in the program might
find group therapy sessions repetitive and unhelpful. Many people with serious
mental disorders—particularly those who are well enough to qualify as
candidates for outpatient (rather than inpatient) commitment—are capable of
forming preferences about treatment that are the product of reasonably rational
reflection on their treatment experiences and goals. Respect for their autonomy
requires respect for their treatment preferences even when others think they are
imprudent.
4. When—If Ever—Is Paternalism Justified?
a. Retrospective Endorsement Theories
Some scholars contend that involuntary mental health treatment can be
justified on a theory of retrospective endorsement.317 The claim here is that
even when people initially refuse treatment, once they have recovered from an
episode of illness, in retrospect, some will be grateful that they were compelled
to accept treatment at a time when they could not think rationally about their
needs.318
Although many clinicians have patients who feel this way, retrospective
endorsement theories of paternalism quickly run into problems. Liberal
theorists have largely rejected subsequent consent theories of paternalism on
conceptual grounds. We are after an ex ante justification for paternalism; at
best, a subsequent consent theory can only provide a normative justification for
an act that has already taken place.319 Moreover, empirical studies on so-called
“thank you theories” of civil commitment have produced mixed results. A large
randomized controlled trial of outpatient commitment in North Carolina found
316. Monika Edlinger et al., Factors Influencing the Choice of New Generation Antipsychotic Medication in the Treatment of Patients with Schizophrenia, 113 SCHIZOPHRENIA RES. 246, 249-50 (2009).
317. See, e.g., ALAN STONE, MENTAL HEALTH AND THE LAW: A SYSTEM IN TRANSITION 18-19 (1975) (advancing his “thank you” theory of civil commitment).
318. DEBRA A. PINALS & DOUGLAS MOSSMAN, EVALUATION FOR CIVIL COMMITMENT 82 (2012).
319. Douglas Husak, Paternalism and Consent, in THE ETHICS OF CONSENT 107
(Franklin G. Miller & Alan Wertheimer eds., 2009); DONALD VANDEVEER, PATERNALISTIC
INTERVENTIONS: THE MORAL BOUNDS OF BENEVOLENCE 66-69 (1986).
2015] INVOLUNTARY OUTPATIENT COMMITMENT 211
little evidence to support a retrospective endorsement theory.320 Instead, as the
authors note, most study participants did not endorse the benefits of outpatient
treatment either because they did not believe that [outpatient commitment] was
effective, or because they refused to believe that they needed treatment, or
both.321
b. The Soft Paternalist Strategy
I shall adopt a soft paternalist theory of intervention and consider
objections from hard paternalism in Part V. Soft paternalism holds that
government interventions into self-regarding harm are justified when, and only
when, the actions or choices of the person concerned are substantially
nonvoluntary, or when temporary intervention is necessary to determine
whether the person’s choices or actions are substantially nonvoluntary. John
Stuart Mill offers a famous example:
If either a public officer or anyone else saw a person attempting to cross a
bridge which had been ascertained to be unsafe, and there were no time to
warn him of his danger, they might seize him and turn him back without any
real infringement of his liberty; for liberty consists in doing what one desires,
and he does not desire to fall in the river.322
If we see someone about to cross an unsafe bridge, we may pull him out of
harm’s way if there is no time to warn him. According to Mill, since “liberty
consists in doing what one desires,” we do not violate his interests in liberty. In
the same way, Feinberg contends that when governments intervene in self-
regarding harm, the concern should not be with the wisdom or prudence of a
person’s choice, “but rather with whether or not the choice is truly his” own.323
Both philosophers recognize defects of reason as a justification for
intervention into self-regarding harm. For example, the harm principle takes a
strong position against paternalism; 324 Mill, however, hastens to add: “[i]t is,
perhaps, hardly necessary to say that this doctrine is meant to apply only to
human beings in the maturity of their faculties.”325 Similarly, Feinberg
enumerates a list of “voluntariness-vitiating” factors—ignorance, coercion,
drugs, and of course, “derangement.”326 Thus, one liberal answer to the
question posed above is that we may intervene into self-regarding conduct
320. Marvin Swartz et al., Endorsement of A Personal Benefit Among Persons with Severe Mental Illness, 9 PSYCH. PUB. POL’Y & L. 70, 90 (2003).
321. Id. at 75-77.
322. JOHN STUART MILL, ON LIBERTY 12 (Elizabeth Rapport ed., 1978) (emphasis added).
323. FEINBERG, supra note 273, at 12.
324. See MILL, supra note 322, at 12 (“The only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others. His own good, either physical or moral, is not a sufficient warrant.”).
325. Id.
326. FEINBERG, supra note 273, at 12.
212 STANFORD LAW & POLICY REVIEW [Vol. 26:159
when that conduct is rendered substantially nonvoluntary by mental
impairment.
In recent years, the notion that schizophrenia and bipolar disorder can
cause damage to parts of the brain has grown tremendously as a moral
justification for involuntary outpatient commitment.327 The argument is that
damage to certain parts of the brain causes impaired insight and therefore we
have reason to interfere with the treatment decisions of people who have been
diagnosed with these disorders. In Part III.C, I argue that we should reject the
impaired insight argument.
C. Impaired Insight
One might think that outpatient commitment orders are justified by virtue
of the fact that, unlike people with general medical conditions, people with
mental illnesses lack the capacity to make treatment decisions on their own.
Therefore, courts may order outpatient treatment for them. But not all people
with mental illnesses are incompetent to make treatment decisions,328 and most
outpatient commitment laws do not require a determination of incompetence.
Instead supporters of involuntary outpatient commitment, foremost among
them psychiatrist E. Fuller Torrey, argue that outpatient commitment should be
considered for anyone with a severe psychiatric disorder who lacks insight into
his or her illness and is at risk of harming themselves or others.329
1. Impaired Insight Defined
In psychiatry, the term “insight” generally refers to the patient’s awareness
that he or she is suffering from an illness. In an early and influential paper on
the psychopathology of insight, Aubrey Lewis described insight as “a correct
327. See, e.g., 113 CONG. REC. H6749-50 (daily ed. July 24, 2014) (statement of Rep. Murphy) (discussing anosognosia and The Helping Families in Mental Health Crisis Act, H.R. 3717); Kress, supra note 2, at 1274; Torrey & Zdanowicz, supra note 14, at 338; see also Dale, supra note 16 (discussing impaired insight); Guido Zanni et al., The Effectiveness and Ethical Justification of Psychiatric Outpatient Commitment, 7 AM. J. BIOETHICS 31, 33 (2007) (discussing anosognosia as an important justification for outpatient commitment);.
328. See generally Thomas Grisso & Paul S. Appelbaum, The MacArthur Treatment Competence Study III, LAW & HUM. BEHAV. 149, 169-72 (1995) (discussing three measures of competence—understanding, appreciation, and reasoning). The MacArthur Treatment Competence Study found that patients with schizophrenia and depression were more likely to show decisional deficits when compared to patients with heart disease and a non-medically ill control group. However, despite the overall poorer performance of the mentally ill patients, on any given measure of competence, most patients with schizophrenia did not score below patients with heart disease or the non-medically ill control group. Instead the generally poorer performance of people with schizophrenia was attributed to a minority of patients within the group whose symptoms were most severe.
329. E. Fuller Torrey & Mary Zdanowicz, Outpatient Commitment: What, Why and for Whom?, 52 PSYCHIATRIC SERVICES 337 (2001).
2015] INVOLUNTARY OUTPATIENT COMMITMENT 213
attitude toward morbid change in oneself.”330 Modern approaches define
insight along similar dimensions. In a foundational article on insight and
psychosis, Anthony David conceptualized insight along three distinct, though
overlapping dimensions—(i) recognition that one has a mental illness; (ii)
compliance with treatment; and (iii) the ability to relabel unusual mental
events, such as delusions and hallucinations, as pathological.331
Xavier Amador and D.H. Strauss have advanced a similar model in which
insight consists of two components—(i) awareness of illness and (ii) attribution
of illness.332 As Amador and Kronengold write: “[u]nawareness of illness
reflects an individual’s failure to acknowledge the presence of a specific defect
or sign of illness even when confronted with it by an examiner,” while incorrect
attribution, on the other hand, “reflects the individual’s expressed belief that the
specific deficit, sign or consequence of illness does not stem from mental
dysfunction.”333 The widely used Scale to Assess Unawareness of Mental
Disorder (SUMD) based on their model allows interviewers to assess current
and retrospective awareness of having a mental disorder. Interviewers use the
SUMD to assess awareness of particular signs and symptoms of mental
disorder, awareness of treatment benefits, and the psychosocial consequences
of having a disorder.334
2. The Neuroscience of Insight
In the debate surrounding Kendra’s Law, neuroscientific evidence
regarding the biological basis of impaired insight has played an important role
in the moral justification for outpatient commitment. E. Fuller Torrey and Mary
Zdanowicz write:
We argue that outpatient commitment is needed because many individuals
with severe psychiatric illness lack awareness of their illness. This deficit is
biologically based and is not the same thing as psychological denial. Both
schizophrenia and bipolar disorder affect the prefrontal cortex, which is used
for insight and understanding one’s needs. When this area of the brain is
damaged, the person loses self-awareness.335
330. Aubrey Lewis, The Psychopathology of Insight, 14 BRIT. J. PSYCHOL. 332, 333 (1934); see also Anthony S. David, “To See Ourselves as Others See Us: Aubrey Lewsis’s Insight,” 175 BRIT. J. PSYCHOL. 210 (1999).
331. Anthony S. David, Insight and Psychosis, 156 BRIT. J. PSYCHIATRY 798, 805 (1990).
332. Xavier F. Amador & D.H. Strauss, Poor Insight in Schizophrenia, 64 PSYCHIATRIC
Q. 305, 307 (1993).
333. Xavier F. Amador & Henry Kronengold, The Description and Meaning of Insight in Psychosis, in INSIGHT AND PSYCHOSIS: AWARENESS OF ILLNESS IN SCHIZOPHRENIA AND
RELATED DISORDERS 17 (Xavier F. Amador & Anthony S. David eds., 1998).
334. Xavier F. Amador et al., The Assessment of Insight in Psychosis, 150 AM. J. PSYCHIATRY 873 (1993).
335. Torrey & Zdanowicz, supra note 14, at 338.
214 STANFORD LAW & POLICY REVIEW [Vol. 26:159
Torrey and Zdanowicz are outspoken proponents of outpatient commitment
and the directors of the Treatment Advocacy Center, a non-profit organization
dedicated to eliminating barriers associated with the treatment of mental illness
and promoting outpatient commitment laws.336 Torrey and Zdanowicz argue
that impaired illness awareness, common among patients with schizophrenia,
resembles anosognosia among patients with neurological disorders such as
Alzheimer’s disease, or patients who have suffered a stroke.337
In classic cases
of anosognosia, paraplegic patients who have suffered damage to the right
hemisphere of the brain will deny that they are paralyzed on the left side of the
body. When confronted with the affected limb, anosognostic patients
sometimes insist that the limb is not their own or express indifference in
response to their paralysis.338 In the same way, it is not uncommon for patients
with schizophrenia to deny the symptoms of mental illness. A study by Xavier
Amador and colleagues assessed more than 400 patients with psychotic
disorder and found that nearly 60% of patients with schizophrenia were
unaware of having a mental illness. When asked whether they had any mental,
psychiatric, or emotional problems, most patients answered with an emphatic
“no.”339 When compared to patients with bipolar disorder or schizoaffective
disorder, patients with schizophrenia were also less likely to acknowledge
specific symptoms of mental disorder, including delusions, hallucinations,
thought disorder, and blunted affect.
Torrey and Zdanowicz argue that impaired illness awareness is biologically
based and therefore distinguishable from mere psychological denial. For Torrey
and Zdanowicz, the neurobiological basis of impaired illness awareness
furnishes a critical distinction between people with mental illnesses and other
people with general medical conditions who sometimes refuse treatment. As
the authors write, we can assume that when people with heart disease or
arthritis refuse treatment, “their cognitive functioning and awareness of their
illness are intact”; however, “one cannot make this assumption about an
individual who has a severe psychiatric disorder.”340 In the same way, other
supporters of outpatient commitment argue that community treatment orders
are justified for those whose “brain disorders prevent them from making an
informed decision.”341
Although plausible, arguments along these lines are not without their
problems. First, while many studies have found a significant relationship
between impaired insight and poor performance on tests of frontal lobe
336. Who We Are and What We Do, TREATMENT ADVOC. CENTER, http://www. treatmentadvocacycenter.org/about-us (last visited Dec. 14, 2014).
337. Torrey & Zdanowicz, supra note 14, at 337.
338. See Amador & Kronengold, supra note 333, at 26.
339. Xavier F. Amador et al., Awareness of Illness in Schizophrenia, 51 SCHIZOPHRENIA
BULL. 826, 828-29 (1994).
340. Torrey & Zdanowicz, supra note 14, at 338.
341. Dale, supra note 16, at 274 (quoting Mark R. Munetz et al., The Ethics of Mandatory Community Treatment, 31 J. AM. ACAD. PSYCHIATRY & L. 173, 175 (2003)).
2015] INVOLUNTARY OUTPATIENT COMMITMENT 215
function, many studies of comparable design and quality have not.342
Researchers have also used magnetic resonance imaging scans to examine the
relationship between impaired insight and structural abnormalities in the brain,
but here again the findings are inconsistent.343
Instead, the empirical literature suggests that there multiple pathways to
impaired insight, and many of them have little or nothing to do with cognitive
impairment. In some instances, a denial of mental illness may result from
psychological denial regarding the severity of symptoms. Researchers posit that
at least for some patients the pretense of illness unawareness is a coping
mechanism whereby the symptoms of psychosis are recast as positive events or
avoided in order to avoid the stigma of schizophrenia.344 Similarly, a person
might deny that he has a mental illness because he believes that he has been
wrongly diagnosed with a mental illness or at the very least misdiagnosed. For
the same reasons, a person might deny the benefits of treatment or understate
the consequences of refusing treatment because in his experience he has been
overmedicated or inappropriately medicated. We know that African American
men—who are often the subject of court-ordered outpatient treatment—are
significantly more likely to be misdiagnosed with schizophrenia, giving them
good reason to challenge their diagnoses and the benefits of treatment in
court.345
Even when a person is prepared to acknowledge troubling or distressing
symptoms to himself, stigma surrounding the term “mental illness” can
sometimes cause a person to deny that his experience is properly classified as a
mental illness. For others, a denial of mental illness will stem from a
fundamental difference of opinion about what it means to have an “illness.”
Whether a person understands himself as ill will depend on his experience of
what it means to be ill, the meaning of the term “illness” in the world around
him, and his observation of others who have been classified as ill.
342. See generally IVANA S. MARKOVÁ, INSIGHT IN PSYCHIATRY 103 (2005) (reviewing the empirical literature on insight and psychosis).
343. See, e.g., Laura Flashman et al., Smaller Brain Size Associated with Unawareness of Illness in Patients with Schizophrenia. 157 AM. J. PSYCHIATRY 1167 (2000). But see S.L. Rossell et al., Insight: Its Relationship with Cognitive Function, Brain Volume and Symptoms in Schizophrenia, 33 PSYCHOL. MED. 111 (2003) (finding no association between insight and neuroanatomical measures, including white- and grey-matter volume).
344. Michael Startup, Awareness of One’s Own and Others’ Schizophrenic Illness, 26 SCHIZOPHRENIA RES. 203, 203 (1997); see also Paul Lysaker et al., Patterns of Neurocognitive Deficits and Unawareness of Illness in Schizophrenia, 191 J. NERVOUS &
MENTAL DISEASE 38 (2003) (positing that subgroups of patients with poor insight may show poor insight for different reasons including poor cognition and poor reality testing as well as a tendency to deny unpleasant things).
345. See Arnold Barnes, Race and Hospital Diagnoses of Schizophrenia and Mood Disorders, 53 SOC. WORK 77, 80 (2008). See generally JONATHAN MENZEL, THE PROTEST
PSYCHOSIS: HOW SCHIZOPHRENIA BECAME A BLACK DISEASE (2011).
216 STANFORD LAW & POLICY REVIEW [Vol. 26:159
3. Impaired Insight in The Courtroom
Evidence regarding impaired insight can also play a role in assisted
outpatient commitment hearings.346 For example, although Kendra’s Law does
not use the term “insight,” court-ordered outpatient treatment requires evidence
that “as a result of his or her mental illness,” the subject of a petition for AOT
is unlikely to participate in outpatient treatment voluntarily.347 In the same
way, the legislative findings for Kendra’s Law note that there are some people
with mental illnesses who are capable of surviving safely in the community but
often reject voluntary outpatient treatment “because of their illness.”348 To that
end, it is not uncommon for clinicians to reference poor insight as evidence that
the person is unlikely to cooperate with outpatient treatment absent court
ordered supervision.349
Clinical judgments regarding the patient’s level of insight are
extraordinarily difficult to challenge in court. Accounts from patients regarding
their illness are routinely discounted. Judgments about insight may also depend
on the patient’s attitudes toward treatment and whether patients agree with their
diagnosis. For example, in a small study where psychiatrists were asked to
describe their understanding of insight and its relationship to diagnosing
schizophrenia, researchers found that psychiatrists tended to construct insight
as the patients’ ability “to conceptualize what is happening to them in terms of
the dominant model of mental illness.”350 The authors continued: “[n]ot only is
the patient supposed to understand that they are ill, but crucially, they are to
346. See generally Candice T. Player, Outpatient Commitment and Procedural Due Process, 38 INT’L J.L. & PSYCHIATRY (forthcoming 2015) (on file with author) (finding that in a little more than half of the contested cases observed, psychiatrists testified that the re-spondent lacked insight into his or her mental illness or the need for treatment). Researchers in the United States have paid little attention to the role of impaired insight in outpatient commitment proceedings. Observers in England, New Zealand and Australia have, however, questioned the role of insight in the deliberation of mental health tribunals. These tribunals are responsible for issuing community treatment orders and deciding whether to discharge patients. See, e.g., Kate Diesfeld, Insights on “Insight”: The Impact of Extra-Legislative Factors on Decisions to Discharge Detained Patients, in INVOLUNTARY DETENTION AND
THERAPEUTIC JURISPRUDENCE: INTERNATIONAL PERSPECTIVES ON CIVIL COMMITMENT (Kate Diesfeld & Ian R. Freckelton eds., 2003); John Dawson & Richard Mullen, Insight and Use of Community Treatment Orders, 17 J. MENTAL HEALTH 269, 278 (2008) (finding that impaired insight was a reason for continuing treatment orders in a qualitative study of forty-two outpatients, clinicians, and caregivers in New Zealand).
347. N.Y. MENTAL HYG. LAW § 9.60(c)(5) (2006) (McKinney, Westlaw through 2014 Sess.).
348. Mental Hygiene—Community-Based Settings—Kendra's Law, ch. 408, S.5762-A, 1999 N.Y. Sess. Law (codified as amended at N.Y. MENTAL HYG. LAW § 9.60 (McKinney, Westlaw through 2014 Sess.).
349. See generally Player, supra note 346.
350. Dariusz Galasinski & Konrad Opalinski, Psychiatrists’ Accounts of Insight, 22 QUALITATIVE HEALTH RES. 1460, 1463 (2012).
2015] INVOLUNTARY OUTPATIENT COMMITMENT 217
adopt and accept the psychiatric view and understanding of their
experiences.”351
An attorney who represents respondents in assisted outpatient treatment
proceedings expressed a similar frustration. When her clients are asked whether
they have a mental illness, she explained that “doctors are looking for a very
specific answer to that question” even though many of her clients would prefer
to describe their symptoms as the product of a “chemical imbalance” or
anything other than a mental illness.352 The same attorney wondered whether
her clients would ever be able to verbalize the symptoms of their mental
illnesses in a way that would convince the court of their insightfulness.
It may be that we underestimate what it would require to enter a
courtroom, sit before a judge, and make an oral argument on one’s own behalf.
The ideal respondent would describe her symptoms in detail, eloquently
remarking on the textbook definition of her illness and her particular illness
experience. However, when respondents fall short of this standard, what are we
really hearing? How much of their lackluster testimony is the result of poor
education, lack of preparation, anxiety, or the foreignness of the forum and the
foreignness of the task itself?
The attorney felt that what should matter in these proceedings is whether
her clients are actually complying with treatment. However, even if insight
were a function of the respondent’s behavior, it would be incredibly difficult
for the respondent to convince the court that a period of treatment compliance
during the preceding AOT order was the result of greater insight. The problem
for trial judge is plain—how can a judge know whether a period of treatment
compliance was the product of greater insight or the threat of a seventy-two
hour hospitalization?
4. An Objection from Autonomy
Thus far, I have argued that we ought to reject impaired insight as a
morally significant distinction between people with mental illnesses and others
on largely epistemic grounds. However, impaired insight approaches are also
vulnerable to objections from autonomy. To ask whether a person possess
insight into his illness is to ask whether he recognizes his symptoms as the
symptoms of a mental illness, whether he appreciates the seriousness of those
symptoms, and whether he recognizes the need for treatment—in short,
whether he possesses “a correct attitude toward morbid change in oneself.”353
351. Id. at 1465.
352. Interview with Jane Doe, Attorney, Mental Hygiene Legal Service, in New York, N.Y. (June 15, 2011) (on file with author).
353. Aubrey Lewis, The Psychopathology of Insight, 14 BRIT. J. PSYCHOL. 332, 333 (1934); see also Anthony S. David, “To See Ourselves as Others See Us,” 175 BRIT. J. PSYCHOL. 210 (1999).
218 STANFORD LAW & POLICY REVIEW [Vol. 26:159
In his work on paternalism, Joel Feinberg contends that autonomy refers to
a multiplicity of virtues beyond an interest in self-determination, and foremost
among them is a person’s interest in being “his own man,” or “her own
woman,” and importantly, her interest in maintaining a distinct self-identity.354
Feinberg’s conception of autonomy captures much of what a person might find
offensive about a clinical concept that require patients to endorse a biomedical
understanding of their experience. In Part IV, I argue that we should reject
impaired insight as a measure of diminished mental capacity and instead limit
assisted outpatient treatment to people with mental illnesses who are unable to
make competent treatment decisions on their own.
III. THE LIMITS OF PREVENTION
The central concern of this Article can be thought of in terms of the
following question: under what circumstances can we impose substantial
restraints on individual liberty because we believe a person is likely to harm
himself or others before he has actually done so? Let me begin to answer this
question by contrasting consequentialist and non-consequentialist theories of
moral justification.
A utilitarian or consequentialist theory of outpatient commitment would
ask whether, in the balance of benefits against harms, the good associated with
outpatient commitment outweighs the harms associated with infringing
personal autonomy and the right to refuse treatment. For a utilitarian the
balance weighs heavily in favor of outpatient commitment. The harms to be
avoided are grave—chronic homelessness, violent crime, violent victimization,
incarceration, and suicide. In Part III we saw that although the evidence on
effectiveness is mixed, at least under some circumstances, outpatient treatment
orders are associated with substantial welfare gains—fewer hospitalizations, a
greater likelihood of receiving appropriate medications, and fewer arrests. If we
adopt the view of the Court of Appeals in Matter of K.L., the interference with
personal autonomy is minimal. Therefore we are unlikely to create more harms
than we prevent by requiring people with mental illnesses to participate in
outpatient treatment programs.
At the same time, utilitarian theories violate many of our intuitions about
the special respect owed to persons.355 From a utilitarian outlook, rights have
354. FEINBERG, supra note 273, at 32.
355. See JOHN RAWLS, A THEORY OF JUSTICE 3-4 (1971). Rawls observes: Each person possesses an inviolability founded on justice that even the welfare of society as a whole cannot override. For this reason justice denies that the loss of freedom by some is made right by the greater good shared by others. Therefore in a just society the liberties of equal citizenship are taken as settled; the rights secured by justice are not subject to . . . the calculus of social interests.
Id. Here Rawls refers eloquently to “the liberties of equal citizenship.” Id. Later in Political Liberalism, it becomes clear that he has in mind the liberties secured by the federal Constitution, including rights to privacy, secured by the Due Process Clause, and by extension—I argue—the liberty interest in refusing unwanted medical treatment. See JOHN
2015] INVOLUNTARY OUTPATIENT COMMITMENT 219
no moral force aside from their contribution to utility. Yet rights protect myriad
interests in autonomy and reflect deeply held moral norms. Consider our earlier
conceptions of autonomy and the right of a competent person to refuse
treatment. For the person whose competence is being assessed, the right to
accept or refuse treatment will often hold a combination of instrumental and
non-instrumental value. Although the physician brings her knowledge of
medicine and health to the doctor patient relationship, “health is only one value
among many.”356 Once the physician has informed me of the risks and benefits
associated with a particular treatment, I am in the best position to determine
“which intervention, if any, best serves my wellbeing, as I conceive it.”357 Thus
the right to make important treatment decisions may be instrumentally valuable
in the promotion of wellbeing.
However, even when others are in a better position to make choices for us,
most people want to make important choices about their own lives. Often our
choices, ranging from the mundane and everyday, to important choices about
how to live our lives, have both instrumental and non-instrumental value. For
example, choices about medical treatment are likely to have considerable
symbolic value.358 Most people who are competent to make decisions about
medical treatment are permitted to do so. Therefore, I may value choice
because without it, I will feel that the absence of choice is degrading. If I am a
person of faith, I may value choices regarding medical treatment—including
the choice to forego treatment—because it matters to me that I am able to
incorporate elements of my faith into healing. In these moments, my choices
about treatment may have considerable demonstrative value.
Part IV develops a nonconsequentialist or rights-based theory of outpatient
commitment. The reasons for adopting a rights-based theory are intimately
connected to our view of human beings as moral persons who are capable of
autonomy, and indeed persons with many interests in autonomy.359 To advance
a nonconsequentialist theory of outpatient commitment is not to say that
consequences do not matter, but rather that given our view of persons, moral
rights to non-interference cannot be infringed simply because doing so would
produce gains in utility, social welfare, or even individual wellbeing. Part IV.A
begins with self-regarding harm. When our primary concern is one of self-
regarding harm, I argue that a court order to participate in outpatient treatment
RAWLS, POLITICAL LIBERALISM 227-28 (1993); see also DWORKIN, supra note 18, at 171-83 (1977) (rejecting utilitarianism as inconsistent with the inconsistent with rights of equal concern and respect owed to persons).
356. ALLEN E. BUCHANAN & DAN W. BROCK, DECIDING FOR OTHERS: THE ETHICS OF
SURROGATE DECISION MAKING 30 (1990).
357. Id.
358. See T.M. Scanlon, The Tanner Lectures on Human Values at Brasenose College, Oxford University: The Significance of Choice (May 16, 23 & 28, 1986) (discussing the symbolic and demonstrative value of choice).
359. Dan W. Brock, Involuntary Civil Commitment: The Moral Issues, in MENTAL
ILLNESS: LAW AND PUBLIC POLICY 147, 161 (Baruch A. Brody & H. Tristram Engelhardt, Jr. eds., 1980).
220 STANFORD LAW & POLICY REVIEW [Vol. 26:159
may be appropriate, but only for people with mental illnesses who are
incompetent to make treatment decisions on their own.
In Part IV.B, I turn to the dilemma of other-regarding harm. Moral rights
reflect a view of persons as moral agents who are not only capable of autonomy
but moral responsibility. Although there is a large literature on moral
responsibility, I shall rely on a theory of moral responsibility as responsiveness
to reasons.360 By moral responsibility, I mean the ability to recognize both
moral and prudential reasons for or against an action, the ability to understand
how reasons fit with actions, and the ability to use reason to guide our
actions.361
A pre-crime system where agents identify and seize persons for crimes
they are predicted to commit might produce gains in utility. Yet because we
view human beings as unique in their capacities for moral agency, in a liberal
society, the criminal process is—and must be—the primary mechanism by
which governments deprive citizens of their liberty when they threaten harm to
others. For reasons similar to those articulated by the Supreme Court in Kansas
v. Hendricks and Kansas v. Crane, Part IV.B asserts that the moral legitimacy
of outpatient commitment depends on a more explicit and principled limitation
to people who are dangerous but also lacking in the moral capacities for
criminal responsibility than is currently the case under outpatient commitment
law.
A. Harm to Self
1. Competence to Refuse Treatment
Involuntary outpatient commitment may be appropriate for people with
mental illnesses who are unable to make competent treatment decisions on their
own. But what does competence entail? As Alan Buchanan and Dan Brock
have observed, competence is always competence “to do something” at a
particular time, under particular circumstances; therefore, the appropriate
concept of competence is decision-relative and variable.362 Second, settling on
an appropriate competence standard is not simply a matter of settling on the
correct test, but rather a process of balancing competing values and guarding
against two kinds of error.363 The first error (Type I or false positive) results
from choosing a standard of competence that is too low and failing to protect
the person from the harmful consequences of his or her decisions when those
decisions stem from serious defects in the capacity to decide. The second error
360. I rely on Fischer and Ravizza’s account of moral responsibility as responsiveness to reasons. JOHN MARTIN FISCHER & MARK RAVIZZA, RESPONSIBILITY AND CONTROL: A
THEORY OF MORAL RESPONSIBILITY (1998).
361. Id. at 69.
362. BUCHANAN & BROCK, supra note 356, at 18.
363. Id. at 40-41.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 221
(Type II or false negative) results when we choose a threshold for competence
that is too high and fail to allow a person to make her own choices when she is
able to do so.364
Like most authors on competence, I agree that competence to refuse
treatment requires the ability to understand key facts involved in a decision to
refuse treatment, the ability to engage in basic reasoning about those facts, the
ability to reach a decision, and the ability to communicate a stable choice.
However, in contrast to other authors on competence, I argue that an emphasis
on appreciation or insight as a measure of competence is misplaced.365 Indeed
“appreciation,” the legal correlate of insight, should have no role to play in our
thinking about competence.
a. Understanding and Appreciation
A person who is competent to refuse treatment must possess at least a
rudimentary understanding of the basic features of his illness and the proposed
treatment plan. Whether he believes he has an illness or not, and whether he
believes that treatment will help him or not, he must at least understand that his
physician believes he has an illness and that his physician believes the
recommended treatment could help him. Any less, and we would worry that the
person is too impaired or too disoriented for us to view his treatment decisions
as competent.
As to this element of competence, it should be enough that the person
understands his decision in this basic factual sense—he is aware of his medical
diagnosis, he can explain it in lay terms, and he understands the proposed
treatment plan, as well as the primary risks and benefits associated with
treatment. However, most authors on competence think that understanding
should incorporate a notion of understanding as appreciation or insight.366 For
364. Id.
365. See, e.g., THOMAS GRISSO & PAUL S. APPELBAUM, ASSESSING COMPETENCE TO
CONSENT TO TREATMENT: A GUIDE FOR PHYSICIANS AND OTHER HEALTH PROFESSIONALS 31 (1998) (recommending four functional abilities as the focus for assessments of competence to consent to treatment, including “the ability to appreciate the significance of information for one’s own situation, especially concerning one’s illness and the probable consequences of one’s treatment options”); see also Kathleen Glass, Refining Definitions and Devising Instruments: Two Decades of Assessing Mental Competence, 20 INT’L J.L. & PSYCHIATRY 5, 22 (proposing that “appreciation implies sufficient critical judgment or insight to value the information that has been comprehend” and when emotional or affective factors inhibit the formation a judgment regarding mental health treatment, that assessment should result in a finding of incompetence); Louis C. Charland, Mental Competence and Value: The Problem of Normativity in the Assessment of Decision-Making Capacity, 8 PSYCHIATRY, PSYCHOL. &
L. 135, 136 (2001). But see Christopher Slobogin, Appreciation as a Measure of Competency: Some Thoughts About the MacArthur Group’s Approach, 2 PSYCHOL. PUB. POL’Y & L. 18 (1996).
366. See, e.g., Grisso & Appelbaum, supra note 328, at 155-56 (1995) (discussing subtests of the MacArthur Competence Assessment Tool (MacCAT)—the Nonacknowledgement of Disorder subtest (NOD) and the Nonacknowledgement of
222 STANFORD LAW & POLICY REVIEW [Vol. 26:159
example, Buchanan and Brock contend that understanding requires the ability
to “appreciate the nature and meaning of potential alternatives—what it would
be like and ‘feel’ like to be in possible future states and to undergo various
experiences—and to integrate this appreciation into one’s decision making.”367
In doing so, Buchanan and Brock invoke a weak notion of appreciation, one
that requires no more than the ability to imagine what it would be like to
experience a future health state.
Nonetheless, a stronger form of appreciation, advanced by Paul
Appelbaum and Thomas Grisso, includes a notion of appreciation as agreement
or acknowledgment of mental disorder. In the MacArthur Treatment
Competence Study, Appelbaum and Grisso assert that competence consists of
four abilities: the ability to communicate a choice; the ability to understand
relevant information; the ability to manipulate information rationally; and
importantly, the ability to appreciate the nature of the situation and its likely
consequences.368 Appelbaum and Grisso define appreciation as
“acknowledgment of illness and the potential value of treatment.”369 The
authors concede that a person might disavow his diagnosis or fail to
acknowledge the potential value of treatment for reasons other than a
psychiatric disorder. For example, a person might deny the potential value of
treatment owing to religious reasons or a history of unsuccessful
interventions.370 Therefore, the authors contend that nonacknowledgment of
disorder, or the consequences of non-treatment, should only qualify as a failure
of appreciation when patients meet three criteria. First, the patient’s belief must
be “substantially irrational, unrealistic, or a considerable distortion of reality.”
The belief must also be “the consequences of impaired cognition or affect,” and
finally, the belief must be “relevant to the patient’s treatment decision.”371
The MacArthur Treatment Competence Study designed the Perceptions of
Disorder test (POD) to measure appreciation. POD consists of two subtests—
the Nonacknowledgment of Disorder test (NOD) and the Nonacknowledgment
Treatment subtest (NOT)). While the NOD is designed to assess the extent to which patients acknowledge the existence of their mental disorder, the NOT allows patients to rate their degree of agreement or disagreement in response to statements about their disorder and the potential benefits of treatment. Id.
367. BUCHANAN & BROCK, supra note 356, at 24; see also Charles M. Culver & Bernard Gert, Inadequacy of Incompetence, 68 MILBANK Q. 619 (1990) (arguing that a person is competent to decide whether to refuse treatment when she understands information relevant to making a treatment decision and “appreciate[s] how this information applies to oneself in one’s current situation”).
368. Paul Appelbaum & Thomas Grisso, The MacArthur Treatment Competence Study I: Mental Illness and Competence to Consent to Treatment, 19 LAW & HUM. BEHAV. 105, 109-11 (1995).
369. Thomas Grisso & Paul Appelbaum, The MacArthur Treatment Competence Study II: Measures of Abilities Related to Competence to Consent to Treatment, 19 LAW & HUM. BEHAV. 127, 128 (1995).
370. Id. at 132.
371. GRISSO & APPELBAUM, supra note 365, at 45-48.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 223
of Treatment Potential test (NOT).372 NOD assesses the extent to which
patients acknowledge their mental disorder as well as their symptoms; their
beliefs about the severity of their symptoms; and their ability to acknowledge
the diagnosis in their hospital chart.373 Patients’ beliefs about the severity of
their symptoms are compared to their scores on the Brief Psychiatric Rating
Scale (BPRS). Patients are given a score of “0” when they rate their symptoms
as not severe when their symptoms are considered severe according to the Brief
Psychiatric Rating Scale.
The Nonacknowledgment of Treatment subscale (NOT) assesses the extent
to which patients acknowledge that treatment might help them, including their
ability to acknowledge the relevance of treatment; the potential benefit of a
particular treatment; and the likelihood of not improving without treatment.374
Patients who fail to acknowledge the potential value of treatment are asked to
explain why and presented with a hypothetical. For example, when a patient
believes that treatment will not help her because she is simply too sick, the
examiner might respond: “Imagine that a doctor tells you that there is a
treatment that has been shown in research to help 90% of people with problems
just as serious as yours. Do you think this treatment might be of more benefit
to you than getting no treatment at all?”375 Patients who “rigidly disavow” the
value of treatment by indicating definitely or probably not on a six-point scale,
in response to that hypothetical receive a score of “0” on that item.376
In accordance with Grisso and Appelbaum, philosopher Louis Charland
has also endorsed a strong form of appreciation as a necessary element of
competence.377 He writes: “Appreciation consists in an individual’s ability to
apply his or her current understanding of a given medical condition to him or
herself. It is one thing to understand what schizophrenia is, but quite another to
recognise that this information applies to you.”378
Charland and I are in agreement here, but only in a weak sense. In order to
meet the understanding prong of a competence test, it should be enough for a
person to understand that a psychiatric assessment is an assessment of him, not
some hypothetical person. Charland, however, intends something more. By
“recognise,” I take Charland to mean “agree” so that a person must agree that
he has an illness called schizophrenia. However, on the view of competence I
am proposing, agreement on diagnosis would require too much.
Approaches to competence that require some degree of “appreciation” will
fail to account for instances of reasonable disagreement. A reasonable person
might conclude that a particular treatment is “probably” or “definitely” unlikely
372. Grisso & Appelbaum, supra note 369, at 132.
373. Id.
374. Id.
375. Id. at 133.
376. Id.
377. Charland, supra note 365, at 143.
378. Id. at 136.
224 STANFORD LAW & POLICY REVIEW [Vol. 26:159
to help her, even if research has shown that treatment to help ninety percent of
people with problems as serious as her own. The respondent might be
pessimistic, though not pathologically so, or she might simply have good
reason to believe that she has more in common with people who did not benefit
from treatment. Appelbaum and Grisso acknowledge that a person might deny
the potential value of treatment owing to a history of unsuccessful
interventions. Still the Perception of Disorder Test (POD) would classify this
disagreement as a rigid disavowal and thus a failure of appreciation.
Similarly, clinicians and patients may agree on most aspects of a case yet
fail to agree on a diagnosis. A person might acknowledge feelings of sadness,
fatigue, and loss of energy for most of the day, for more than two weeks—and
indeed, acknowledge the recurrence of these feelings over time—yet express
some ambivalence about whether he is experiencing a clinical depression.
Clinicians and patients will sometimes disagree over whether feelings of
sadness and depression are pathological or merely a normal reaction to external
circumstances. Others may feel that their symptoms do not rise to the level of
an illness.
Or consider a person who has been diagnosed with schizophrenia.
Schizophrenia is a psychotic disorder characterized by hallucinations,
delusions, disorganized speech, and disorganized behavior. Although
symptoms of psychosis frequently accompany schizophrenia, a variety of
medical conditions can include psychotic symptoms, including substance-
induced psychosis, delusional disorder, and bipolar disorder.379 The positive
symptoms of schizophrenia—delusions, hallucinations, and paranoia—
resemble the symptoms of mania while the negative symptoms of
schizophrenia—flattened affect, emotional withdrawal, and social isolation—
can resemble depression, leading clinicians to confuse schizophrenia with
bipolar disorder or vice versa.380
As I noted above, African American men tend to be overrepresented
among patients who have been diagnosed with schizophrenia and
underrepresented among patients who have been diagnosed with bipolar
disorder and depression.381 However, large epidemiological surveys designed
to measure the prevalence of mental disorders have shown that the prevalence
of schizophrenia, bipolar disorder, and depression does not vary by ethnicity.
The clinical tendency to overdiagnose schizophrenia among African Americans
might arise for any number of reasons, ranging from cultural differences in the
expression of symptomatology to the cultural distance between clinician and
379. AM. PSYCHIATRIC ASS’N, DIAGNOSTIC AND STATISTICAL MANUAL OF MENTAL
DISORDERS 297-98 (Michael B. First ed., 2000).
380. Frederic C. Blow et al., Ethnicity and Diagnostic Patterns in Veterans with Psychoses, 39 SOC. PSYCHIATRY EPIDEMIOLOGY 841, 842 (2004).
381. See supra note 345 and accompanying text.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 225
patient.382 The result, however, is that a reasonable person might “recognise”
that a diagnosis of schizophrenia applies to him yet insist that he has been
wrongly diagnosed nonetheless.
b. Reasoning and Communication
Competence also requires an ability to reason about treatment and the
ability to communicate a choice. What is required here is not perfect rationality
but rather at least a basic ability to reach conclusions that are logically
consistent with major and minor premises.383 Doing so will require an ability to
weigh the risks and benefits of treatment against one’s values as well as at least
a basic understanding of probabilities. For example, a person who is competent
to refuse a recommended course of treatment should understand what it would
mean for an outcome to be more likely than not.384
Most authors on competence agree that an assessment of competence
should focus on the quality of the reasoning process rather than the rationality
or reasonableness of the outcome.385 Our focus on the quality of the reasoning
process should also include limits on the kinds of reasons that are permitted to
factor into a competent decision. The problem, of course, lies in determining
which reasons to exclude. Some reasons are irrelevant as obvious non-sequiturs
and provide ready grounds for exclusion: I am refusing treatment because today
is Tuesday.386 Other reasons are, to use Elyn Saks’ term, “patently false”:
Zyprexa is made of green cheese; my psychiatrist is an alien.387
Saks defines patently false beliefs as beliefs that are “so unlikely that even
the most superficial reading of the data will indicate their falsity.”388 Patently
false beliefs are also distinguishable from simple delusions. To use her
example, consider a patient who suffers from depression and thinks of herself
as a bad person, although others think she is a very good person.389 Her belief
may be false and a delusion, but it is not patently or indisputably false. If, on
the other hand, she were to believe that she is an evil person because she
committed mass murder, even though she had not, such a belief would be
patently and demonstrably false. When a person harbors patently false beliefs
382. Lloyd H. Rogler, Framing Research on Culture in Psychiatric Diagnosis: The Case of the DSM-IV, 59 PSYCHIATRY: INTERPERSONAL & BIOLOGICAL PROCESSES 145, 147
(1996).
383. Paul Appelbaum & Thomas Grisso, Assessing Patients’ Capacities to Consent to Treatment, 319 NEW ENG. J. MED. 1635, 1636 (1988).
384. Id.
385. See, e.g., Charles M. Culver & Bernard Gert, The Inadequacy of Competence, 68 MILBANK Q. 619, 621 (1990); Benjamin Freedman, Competence: Marginal and Otherwise, 4 INT’L J.L. & PSYCHIATRY 53, 53 (1981).
386. See Freedman, supra note 385, at 64.
387. Elyn Saks, Competency to Refuse Treatment, 69 N.C. L. REV. 945, 956 (1991).
388. ELYN SAKS, REFUSING CARE: FORCED TREATMENT AND THE RIGHTS OF THE
MENTALLY ILL 187 (2002).
389. Id. at 186-87.
226 STANFORD LAW & POLICY REVIEW [Vol. 26:159
despite evidence to the contrary, we have good reason to believe that he or she
is incompetent to make treatment decisions.
Still, not all of the reasons one might offer for refusing treatment are
obviously irrelevant or susceptible to ready proof as patently and demonstrably
false. In the first category are factual questions that are susceptible to proof. In
the second category are reasons based on beliefs, the most challenging of which
may be reasons for treatment refusal based on religious beliefs. Consider the
Jehovah’s Witness who refuses a blood transfusion owing to his belief that the
Bible prohibits ingesting blood and contains the word of God. We can neither
prove nor disprove the existence of God, yet our commitment to religious
pluralism necessitates some allowance for reasons of this kind to factor into
treatment decisions.
And yet if we accept faith as a legitimate reason for treatment refusal in
these circumstances, how can we distinguish the Jehovah’s Witness from the
person who refuses treatment, not owing to his belief in God, but owing to his
belief that he is God? The Jesus delusion is the most common of all psychiatric
delusions, combining both delusions of grandeur and persecution.390 Clinicians
frequently encounter patients who believe they are Jesus, the Virgin Mary, or
other figures of religious or historical importance, and a claim to be any one of
them is virtually certain to result in a diagnosis of serious psychopathology.391
For some, the Jesus delusion results from a confusion of similarities and
identities—Jesus was a man with a blondish beard; I am a man with a blondish
beard; therefore, I am Jesus.392 We can imagine circumstances in which
persons come to see themselves as Jesus in a merely metaphorical sense that
can be clarified in conversation. In these circumstances, the Jesus delusion by
itself would not be sufficient grounds to conclude that a person is incompetent
to make treatment decisions. However, most of us probably want to say that a
person who persists in his belief that he is in fact the risen son of God—and
indeed refuses treatment for that reason—is not competent to make treatment
decisions.
When confronted with this question, clinicians will ordinarily ask whether
the patient’s beliefs predate the treatment decision, whether the patient has
previously behaved in ways that are consistent with those beliefs, and whether
the patient’s beliefs are reflective of religious views held by others.393 Insofar
as we are interested in determining whether the patient’s reasons for refusing
treatment reflect genuine religious beliefs, the thought here is that if the beliefs
are unconventional or idiosyncratic to the patient, they are more likely to
indicate psychopathology.
390. Alan Gettis, The Jesus Delusion: A Theoretical and Phenomenological Look, 26 J. RELIGION & HEALTH 131, 131 (1997).
391. Id. at 132.
392. Id.
393. GRISSO & APPELBAUM, supra note 365, at 48.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 227
However, in contrast to the dominant medical view, Elyn Saks argues that
a normatively desirable standard would afford considerable protection to
unconventional or idiosyncratic beliefs. Saks writes that while there may be
some limits on what patients can believe, “limits that are too stringent severely
curtail that patient’s freedom to be unconventional in their pursuit of truth.”394
Since many people hold distorted beliefs, we risk “discriminating against the
mentally ill if we disable them based on their distortions.”395 Instead Saks
draws the line at patently false beliefs, asserting that only patently false beliefs
should disqualify a person from competence.
How should we think about this? Consider the following case study,
involving Ricardo Jesus B.:
Several months before he came to us with a diagnosis of paranoid
psychosis, Ricardo Jesus experienced a series of severe epileptic seizures.
When the seizures ended, Ricardo developed a psychotic condition. His
delusion consisted of the following—during the seizure, epileptic Ricardo died
and Jesus B. survived. He, Ricardo, now Jesus B., had been in heaven seated
at the right hand of the Father. The sick person, Ricardo who everyone in the
village laughed at is dead. In his place Jesus B. survives. Jesus B. is not
identical to Jesus Christ, but close to him since they bear the same name. In
the hospital, Ricardo was always in a good mood and respected by other
patients. One day he asked for a certificate of discharge so politely and with
such irreproachable behavior, that we let him leave. He came back to us
twice—always glad, emanating a naïve saintliness and telling anyone who
wanted to hear him how happy he was to no longer be an epileptic at whom
everyone laughed, and how happy he was to spread the good news to mortals,
first to his neighbors in the little mountain village where he cares for his goats,
and then to anyone else who is willing to receive it.396
The case study comes from Ottor Doerr and Óscar Velásquez in the
Department of Psychiatry, University of Chile. Most of us would probably say
that the doctors who released Ricardo B. were right to do so. With little or no
risk of harm to himself, there is no reason to keep him in a hospital.
Suppose, however, that instead of spreading the good news to his
neighbors, in a little mountain village in Chile, Ricardo B. now lives in New
York City. Call him Ricardo C. He continues to believe that at least in some
way he has been reincarnated as the son of God. Only now, his condition has
deteriorated, and when offered the opportunity to participate in treatment, he
asserts, “God does not take medication!”397 As a result, his condition
deteriorates. Like Joyce Brown before him, Ricardo C. sleeps near hot air vents
and in abandoned buildings for warmth. Although he has no money, no home,
and nowhere to go, he manages to eke out a meager existence through charity.
If we accept Elyn Saks view, then despite a diagnosis of psychosis and the
394. SAKS, supra note 388, at 182.
395. Id.
396. Otto Doerr & Óscar Velásquez, The Encounter with God in Myth and Madness, 2:12 PHIL. ETHICS & HUMAN. MED. 1, 5 (2007).
397. See TORREY, supra note 270, at 116.
228 STANFORD LAW & POLICY REVIEW [Vol. 26:159
obvious risk of serious harms to himself, courts would have no authority to
intervene. His sincere belief that he is the son of God is neither irrelevant nor
patently and demonstrably false; therefore, her standard imposes an absolute
bar against intervention. By drawing the line at patently false beliefs, Saks
protects the right to hold idiosyncratic ideas. However, in doing so she also
overvalues the interest in autonomy.
Earlier I said that a conception of autonomy as authenticity has obvious
purchase in a discussion of mental illness. When a person refuses treatment for
such unusual religious reasons, we will wonder whether their preferences are
truly their own or whether their preferences are a symptom of psychopathology.
For a soft paternalist, the important question in these cases is whether the
person’s conduct is substantially nonvoluntary. The “patently and demonstrably
false” standard advanced by Saks would impose an absolute bar against
intervention, but for Feinberg, the important question is whether a person’s
choices are “voluntary enough.”398 Feinberg proposes a variable standard of
voluntariness: the riskier the conduct and more irrevocable the harm, the
greater the degree of voluntariness that should be required if the person’s
conduct is to be permitted.399
Our concern for Ricardo C. stems primarily from the serious risk of
morbidity and mortality associated with chronic homelessness. Exposure to
extreme weather conditions, untreated medical illnesses, infection, and
insufficient nutrition will often work together to increase the risk of death. In
circumstances like this one, where the risks are sufficiently grave—and a
person’s capacity to make rational choices is sufficiently in question—a
reasonable court might risk a Type II error and order outpatient commitment,
including antipsychotic medication, in an effort to restore the person’s ability to
be self-governing. Reasonable minds might disagree. Perhaps under different
circumstances, such as warmer weather or a city with a robust system of
voluntary mental health services and dedicated outreach workers, a court might
deny a petition for outpatient commitment and instead risk a Type I error if
there are a robust safety net and private citizens who are willing to intervene.
Yet, like the bridge pedestrian, if a court were to grant a petition for
outpatient commitment, once Ricardo C. regains the capacity to be self-
governing, we must let him go. As Mill writes, the possibility that he might
harm himself supplies good reason for “remonstrating with him,” but not for
“compelling him or visiting him with evil in case he do otherwise.”400
2. Competence Without Insight
Above I proposed that we reject impaired insight as a measure of
diminished mental capacity, so that people are free to make their own decisions
398. FEINBERG, supra note 273, at 118.
399. Id. at 118-20.
400. MILL, supra note 322, at 9.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 229
about medical treatment unless they are incompetent to do so, without reference
to appreciation. By focusing on the patient’s recognition that he has an illness,
the patient’s ability to relabel unusual events as pathological, and compliance
with medical treatment, an impaired insight standard simply asks the wrong
question. When our primary concern is one of self-regarding harm, the central
question in an outpatient commitment hearing should not be—do you agree
with Doctor X about the causal origins of your symptoms and the need for
treatment—as it would be if impaired insight were a measure of diminished
capacity. Rather, the central question should be: “Do you understand that
Doctor X believes that some of your thoughts and behaviors are attributable to
Disease Y? Do you understand that according to Doctor X, consequences A, B,
C, and D are likely to follow if you refuse the recommended course of
treatment?” If the person understands the basic facts of his or her illness in this
sense, then provided that his or reasons are at least neither irrelevant nor
patently and demonstrably false, he or she is competent to make decisions
regarding outpatient treatment and courts should not order outpatient treatment
over his or her objection.
Focusing on whether the respondent is competent to refuse treatment has
certain virtues when compared to the current regime. Whether the respondent
truly appreciates the seriousness of his symptoms and the potential value of
treatment can be difficult to discern, thereby increasing the risk of Type II
errors and needless intrusions on autonomy. In the same way questions about
whether the respondent truly appreciates the seriousness of his illness, the need
for treatment rely too heavily on the respondent’s credibility, and in turn
increase the risk of error. As I noted above, when a person has been diagnosed
with a mental illness, their perceptions of their own needs and their own illness
experience are routinely discounted. Reframing the test in terms of competence
would eliminate the need to determine whether the respondent is telling the
truth when he promises to continue treatment voluntarily. Under the proposed
approach, the important question would be whether the respondent is able to
make an important treatment decision on his or her own.
B. Harm to Others
What should we say about a person who presents a substantial risk of harm
to others by virtue of mental illness, but who is competent to refuse treatment
nonetheless?
Following a fight with his mother, during which he “accidentally” pushed
her to the ground, Gary, a 30-year-old man, was admitted to a psychiatric
hospital with a diagnosis of paranoid schizophrenia. According to hospital
records, Gary was “malodorous,” and “experiencing bizarre delusions,”
including a delusional belief that he was growing extra body parts and being
controlled by “Carrie,” who “likes to eat people’s organs with a knife and
fork.” During an inpatient therapy session, Gary threw a psychiatrist against a
wall and struck a resident with his fists, claiming that he was “unable to
control his arms.” After a few weeks in the hospital, the symptoms of
230 STANFORD LAW & POLICY REVIEW [Vol. 26:159
psychosis improve and Gary files a petition to be released from the hospital.
During the hearing, Gary’s mother testifies to Gary’s history of assault,
treatment noncompliance and substance abuse following his release from
psychiatric hospitals. Doctors petition the court for assisted outpatient
treatment in an effort to prevent a relapse of psychosis that would be likely to
result in serious harm to others. Gary, however, refuses to participate in the
program.401
What should we do? Suppose Gary is competent to refuse treatment. He
understands the basic features of his illness and the proposed treatment plan.
By all accounts, his reasons for refusing treatment are neither irrelevant nor
demonstrably false, and he is able to communicate a stable choice. Still, a fair
outpatient commitment program could order Gary to participate in outpatient
treatment, notwithstanding a finding of competence. Our challenge, however,
will be to distinguish Gary—subject to preventive outpatient commitment—
from others whose dangerous behaviors are more appropriately addressed
through the criminal justice system.
To see the need for justification, consider the garden-variety recidivist. Call
him Joe. Suppose Joe also has a long history of violent crime. Like Gary, the
state is able to establish that without supervision—perhaps a supervised living
arrangement or an anger management program—Joe is also very likely to harm
others. What can we do? If Joe were to harass, stalk, or threaten a particular
person, a court might issue a temporary restraining order. However, without
clear and convincing evidence of a substantial threat against a particular person,
courts will not impose limits on Joe’s freedom in order to prevent the very
serious crimes that he is likely to commit. Why? The underlying assumption of
the criminal law is that most people understand the difference between right
and wrong, and most people are able to conform their behavior to the
requirements of the law.
In both Kansas v. Hendricks and Kansas v. Crane, the Supreme Court
affirmed the deterrent and retributive functions of the criminal justice system as
the preferred approach to handling garden variety criminal conduct. And in
both cases, the Court held that states may use civil commitment to detain sex
offenders beyond the expiration of their sentences when a mental abnormality
makes it “difficult, if not impossible for the person to control his behavior.”402
Yet as I noted above, critics argue that the Court’s impaired control standard is
at best confused and overbroad. Instead police power commitments are only
appropriate for persons who are grossly irrational by virtue of mental illness or,
in essence, “too sick to deserve punishment.”403
Part IV.B develops the intuition that persons who are appropriate
candidates for outpatient commitment resemble persons who do not qualify for
401. EWING, supra note 296, at 114-27 (2008); IN THE MATTER OF DAVID DIX, supra note 97.
402. Kansas v. Crane, 534 U.S. 407, 411 (2002); Kansas v. Hendricks, 521 U.S. 346, 358 (1997).
403. Janus, supra note 258, at 298.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 231
criminal punishment either because they are grossly irrational, as Stephen
Morse argues, or because they are unable to conform their conduct to the
requirements of the law. In American legal thought, the absence of these
capacities provides a rationale for the insanity defense. However, in the
remainder of Part IV, I want to suggest that the absence of these capacities can
also identify persons who are appropriate candidates for outpatient civil
commitment. Consider the following rule based on the Model Penal Code
formulation of the insanity defense and Kendra’s Law: (a) A person who is competent to refuse treatment may be ordered to
participate in an outpatient treatment program if, in view of his
history or current behavior, he is likely to harm others as defined
in section (b) of this article; and
(1) as a result of mental illness, he is unlikely to appreciate
the wrongfulness of his conduct; or
(2) as a result of mental illness, he lacks the capacity to con-
form his conduct to the requirements of the law.
(b) As used in this article the words “likely to harm” shall mean a
substantial risk of serious physical harm as manifested by homici-
dal or other violent behavior.
1. Cognitive Impairment
Under part (a)(1) of the proposed rule, a person who is competent to refuse
treatment may be ordered to participate in outpatient treatment if, in view of his
history or current behavior, he is likely to harm others, and as a result of a
mental illness, he is unlikely to appreciate the wrongfulness of his conduct.
Following the Model Penal Code, we can say that a person may be unlikely to
appreciate the wrongfulness of his conduct if, in the past, he has failed to
apprehend material circumstances as a result of a mental illness or “failed to
apprehend the significance of his actions in some deeper sense.”404 Suppose the
fight that landed Gary in the hospital was not an “accident.” Instead, Gary
pushed his mother to the ground based on a delusional belief that his family
was plotting against him and, indeed, trying to kill him. During an outpatient
commitment hearing, Gary’s sister testifies that her brother’s delusional beliefs
about their family are longstanding and all-encompassing. On several occasions
Gary has choked her and thrown her to the ground. During the worst incident,
he held a knife to her throat. On the stand his sister sobbed: “When I asked him,
Gary, why?” he replied: “You’re the devil. You came here to hurt me. Didn’t
you?” Later Gary testifies that the assault on his sister was sanctioned by God.
Earlier I said that a person may be morally responsible for his actions if the
deliberative process leading to those actions is responsive to moral and
prudential reasons. By moral reasons, I mean reasons arising from the moral
duties that persons in a community owe to one another. Prudential reasons, on
404. MODEL PENAL CODE § 4.01 explanatory note (1981).
232 STANFORD LAW & POLICY REVIEW [Vol. 26:159
the other hand, are the practical or self-interested reasons that persons have for
behaving in accord with moral norms. Gary might recognize his interest in
avoiding prison as a prudential reason to refrain from assaulting his sister. He
might even view her as the kind of creature whose status as a person gives rise
to moral reasons to refrain from assault. Yet insofar as he views the attack on
his sister as divinely sanctioned, he is not appropriately responsive to reasons in
the way the law requires.
If he were to harm his sister, Gary would likely prevail on the cognitive
arm of an insanity defense. Although Gary is legally innocent of a crime, the
same cognitive impairments that exempt him from criminal punishment also
provide a moral justification for a preventive measure. The more difficult
question is this: what constitutes clear and convincing evidence that, as a result
of a mental illness, a person is unlikely to appreciate the wrongfulness of his
conduct? By definition, insanity defenses are backward-looking and driven by
conduct that has already taken place. Although risk assessments are necessarily
fraught, as I mentioned above, a past history of violence is one of the best
predictors of violence. The fact that Gary has a long history of violent assault
connected to his delusional beliefs suggests that without supervised medical
treatment, he is likely to engage in similarly assaultive behavior.
2. Volitional Impairment
Alternatively, under part (a)(2) of the proposed rule, a person who is
competent to refuse treatment may be ordered to participate in outpatient
treatment if he is likely to harm others, and as a result of a mental illness, he
lacks the capacity to conform his conduct to the law. Consistent with the Model
Penal Code formulation, what is required here is not that the person manifests a
total inability to conform his conduct to the law, but only that his impairment is
not insubstantial.405 Consider the following statement from Andrew Goldstein.
When questioned by the police, Goldstein attributed his actions to an
“overwhelming urge.”
I walked to the far end of the platform . . . . As I’m walking I felt a
sensation like something was entering me like a ghost or a spirit or something
like that. While I was walking it fell out of me. When I have the sensation that
something is entering me, I get the urge to push, shove or sidekick. As the
train was coming—it—the feeling disappeared and came back several
times . . . .
As I was standing on the platform, there was a woman standing waiting for
the train. She was facing the incoming train and I was standing behind her. I
got the urge to push, kick or punch . . . .
I feel like an aura, or a sensation like you’re losing control of your motor
systems. And then, you lose control of your senses and everything. And then
you feel like something’s entering you. Like you’re being inhabited. I don’t
405. Id.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 233
know. But—and then, it’s like an overwhelming urge to strike out or push or
punch . . . .406
Stephen Morse has long argued that courts should reject the notion of an
“uncontrollable” urge or any other purported loss of control as a justifiable
predicate for civil commitment.407 Along with Robert Schopp, Morse starts
from the premise that civil commitment amounts to a massive curtailment of
liberty, one that can only be justified by limiting its use to people who are not
morally responsible for their conduct. On the other hand, Morse thinks that
only a defect in the capacity for rationality can work as a coherent non-
responsibility criterion.
Without canvassing all of his arguments, below I want to challenge some
of the more important ones, and in doing so suggest that we have reason to
reconsider the use of a volitional impairment standard for outpatient civil
commitment. To begin, Morse rests his arguments on a thin conception of
moral responsibility. Morse takes the capacity for rationality to be the
defining—and indeed the singular—feature of moral responsibility. But why
should that be the case? In various places, Morse writes that our capacity for
reason distinguishes human beings from the rest of the natural world.408
Moreover, it is our capacity for rationality that explains why, as a general
matter, our society does not confine for dangerousness alone, but instead treats
human actors as moral agents who are capable of evaluating their conduct and
responding to the law’s commands. Yet such a narrow conception of moral
responsibility seems strangely lacking.
Suppose you invite me to a dinner party. Reluctantly, I accept. As dinner
drags on, I twitch, I grimace, and I jerk as I wage a silent battle against the ticks
and pops of Tourette’s Syndrome. I say to myself: “I know I’m a good person. I
know I’m a good person. I won’t mention your husband’s beer belly,” but then,
before I know it, out it slips: “Beer belly! Beer belly! Beer belly! BEER
BELLY!” In my horror I knock over a bottle of wine and stain your new dress.
I fully understand that I have done something wrong by offending my friend’s
husband, but I could not help it. Morse takes the position that a mentally
abnormal cause is merely a cause. “Whether a predisposing factor is produced
by a mental disorder or by some other ‘normal’ or ‘abnormal’ cause makes no
difference to whether the agent is responsible. A cause is just a cause, and
causation per se is not an excuse.”409 It may be that Morse has conflated the
fact that an action is fully attributable to an agent with moral responsibility. The
disruption caused by offending your husband, knocking over a bottle of wine
and staining your new dress is fully attributable to me, but my blameworthiness
406. EWING, supra note 296, at 116.
407. Stephen Morse, A Preference for Liberty: The Case Against Involuntary Commitment of the Mentally Disordered, 70 CAL. L. REV. 54, 59 (1982); see Stephen Morse, Culpability and Control, 142 U. PA. L. REV. 1587, 1608 (1994).
408. E.g., Stephen Morse, Uncontrollable Urges and Irrational People, 88 VA. L. REV. 1025, 1065 (2002).
409. Id. at 1040.
234 STANFORD LAW & POLICY REVIEW [Vol. 26:159
is diminished to the extent that my outburst was caused by a neurological
condition that is beyond my control.410
Morse goes on to argue that control impairments are better understood as
defects in the capacity for rationality. There is certainly a sense in which
Goldstein’s urge to “push, shove or sidekick” stems from a mental abnormality
that we can understand, roughly, as a defect in rationality. On the other hand,
the jurors who convicted him of second degree murder did not think so. In
People v. Andrew Goldstein, Andrew Goldstein pled not guilty by reason of
insanity.411 At trial prosecutors established that Goldstein knew that what he
was doing was wrong and Goldstein conceded as much:
Prosecutor: Well, did you expect that she would go off the platform?
Goldstein: No. No. No. No. I would never push anybody off the tracks.
Prosecutor: Because you know it’s wrong.
Goldstein: Yeah.412
Even if Morse is right, and defenses based on a loss of control really are
better understood as defects in the capacity for rationality, juries are unlikely to
appreciate that subtlety. Instead jurors are more likely to understand their duty
as applying the letter of the law, which means not reading a control defense
into an insanity defense without clear textual support. In New York, the
absence of a volitional impairment standard has clear implications for the
insanity defense; as a result, Andrew Goldstein was found guilty, and indeed
blameworthy, for the death of Kendra Webdale. However, the absence of
volitional element would also have implications for an outpatient commitment
statute. Without it, states would not have the power to reach someone like
Andrew Goldstein.
The difficult question involves determining what would constitute clear
and convincing evidence that, as a result of a mental illness, a person lacks the
capacity to conform his conduct to the law. Once again the Supreme Court
decisions in Hendricks and Crane offer some guidance. After Hendricks and
Crane, state courts were left to determine the evidentiary requirements for
410. To bring the example full circle, contrast my actions to an outburst from my ill-mannered little brother who disrupts the dinner party, not owing to a neurological disorder, but because he is simply bored and ready to go home. Even though we are both fully in our right minds, with no obvious defects in the capacity for rationality, commonsense morality suggests that my outburst, though fully attributable to me, is less blameworthy than that of my brother. Morse might respond that I have made “the fundamental psycholegal error” by conflating causation with excuse. See, e.g., Stephen Morse, Brain Overclaim Syndrome and Criminal Responsibility: A Diagnostic Note, 3 OHIO ST. J. OF CRIM. L. 397, 411 (2006). But my claim here is not that causation itself excuses, but rather that an abnormal cause (though perhaps not a complete excuse) gives us good reason to view conduct as less blameworthy than it would be otherwise.
411. Affirmation and Memorandum for Defendant, People v. Goldstein, No. 00527/99 (N.Y. Sup. Ct. 1999).
412. EWING, supra note 296, at 117.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 235
Crane’s “serious difficulty in controlling behavior” requirement.413 Some state
courts rely heavily on expert testimony to determine whether the defendant has
a serious difficulty controlling his or her behavior, while others rely on a
combination of expert testimony and factual findings. For example in In re
Commitment of W.Z., the New Jersey Supreme Court found that W.Z. had
serious difficulty controlling his behavior based on past acts of sexually
motivated violence and several risk assessments, all but one placing him in a
high-risk category.414
Likewise in United States v. Comstock, the Supreme Court upheld the
Adam Walsh Child Protection and Safety Act, a federal statute that authorizes
civil commitment for sex offenders when their sentences have ended.415
Following Comstock, the Federal Bureau of Prisons issued regulations designed
to guide expert assessments as to whether the defendant has a “serious
difficulty” controlling his or her behavior. Not unlike In re Commitment of
W.Z., the Federal Bureau of Prisons has said that relevant evidence might
include evidence based on a risk assessment, as well as evidence of offending
while under supervision. Relevant evidence might also include engaging in
offenses when likely to get caught, statements of intent to reoffend, or an
admission that the person experiences serious difficulty controlling his or her
behavior.416 Similarly, in outpatient commitment cases, courts might conclude
that a person has a serious difficulty controlling his or her behavior based on
events leading to a prior hospitalization. For example, Andrew Goldstein’s
psychiatric record documents several instances in which Goldstein swung or
punched at others for no apparent reason. And on more than one occasion
before the death of Kendra Webdale, Goldstein complained of being unable to
control his arms.417
IV. OBJECTIONS AND CONCLUSIONS
Preventive outpatient commitment laws test the moral limits of our ability
to intervene in the lives of people with mental disabilities before they harm
themselves or others. I have argued that government interventions into self-
regarding harm and other-regarding harm require distinct moral justifications.
When our primary concern is one of self-regarding harm, I have argued that our
inquiry ought to focus on whether the person is competent to refuse treatment.
Without this limitation outpatient commitment programs fail to respect the
autonomy interests of people with mental disorders. If, however, we are
413. See generally Janine Pierson Construing Crane: Examining How State Courts Have Applied Its Lack-of-Control Standard, 160 U. PA. L. REV. 1527 (2012) (surveying state appellate and supreme court approaches to the Supreme Court mandate in Crane).
414. 773 A.2d 97, 116 (2001).
415. 560 U.S. 126, 150 (2010).
416. Civil Commitment of a Sexually Dangerous Person, 73 Fed. Reg. 70,278, 70,281 (Nov. 20, 2008) (codified at 28 C.F.R § 549.95).
417. IN THE MATTER OF DAVID DIX, supra note 97, at 1.
236 STANFORD LAW & POLICY REVIEW [Vol. 26:159
concerned about harm to others, our inquiry ought to focus on whether the
person possesses the moral capacities for criminal responsibility. By not
limiting outpatient commitment orders to people with mental disorders who are
unable to appreciate the wrongfulness of their conduct or unable to control their
behavior, involuntary outpatient commitment laws intrude on the purview of
the criminal law, a result clearly disfavored by Hendricks and Crane.
An approach of this kind is likely to raise a few objections. The first is that
my approach stems from an overvaluation of autonomy. Autonomy matters, but
hard paternalists will argue that our interest in personal autonomy is not the
only interest that matters. In support of hard paternalism, philosopher Richard
Arneson claims that while the right to autonomy protects an important interest
in self-determination, paternalistic interventions can interfere with that interest
to a greater or lesser degree.418 Therefore “if the consequences of not infringing
the right are sufficiently bad and the interest that the right protects will suffer
only a slight enough degree of frustration, one should in these circumstances
act against the right.”419 Alternatively, if “the good of the individual that is at
stake is enormous,” and the interference in self-determination would be very
slight, an absolute antipaternalism would be “fanaticism.”420 Yet who should
determine whether the interference from paternalism is “very slight?” If the
potential benefits to the individual are indeed “enormous,” and the person
protests nonetheless, the person whose liberty is at stake may feel that the
“interference” is in fact an invasion, even when the intervention seems slight to
others.
Arneson arrives at hard paternalism by way of liberal utilitarianism.
Having rejected a consequentialist justification for outpatient commitment, I
will not belabor the point here. However, a utilitarian moral outlook in these
cases fails to account for our intuitions about the noninstrumental values
protected by upholding the right to refuse treatment. In the same way,
Arneson’s justification for hard paternalism rests on a thin conception of
autonomy. As I have argued, moral rights to autonomy protect more than “a
person’s interest in voluntarily disposing of his lot in life.”421 Insofar as we
view autonomy as a defining feature of personhood, protecting the agent’s
interest in autonomy also protects her interest in viewing herself and being
viewed by others as a creature who is capable of autonomy.
Others have argued for hard paternalism from a public health or
population perspective. The focus of public health is squarely on the health of
populations or communities as a whole, rather than individuals. As Lawrence
Gostin rightly points out, in the aggregate, even conduct that is primarily self-
418. Richard Arenson, Joel Feinberg and the Justification of Hard Paternalism, 11 LEGAL THEORY 259, 263 (2005).
419. Id. at 263.
420. Id. at 264.
421. Id.
2015] INVOLUNTARY OUTPATIENT COMMITMENT 237
regarding can have an adverse impact on social welfare.422 On his view, these
aggregate effects justify some degree of public health paternalism. Nonetheless,
just as Arneson limited the case for hard paternalism to instances where the
interference with autonomy is “very slight,” Gostin argues for hard paternalism
when the intervention does not impose “a truly significant burden on individual
liberty.”423 Neither is willing to surrender important individual interests to a
utilitarian moral calculus or the welfare of the community as a whole.
Rights have costs, among them morbidity and mortality. An approach that
restricts involuntary outpatient commitment to those who are incompetent to
make treatment decisions or otherwise incompetent to bear the burdens of the
criminal law would place many people with serious mental disorders beyond
the scope of court ordered treatment. Yet respect for personal autonomy entails
respect for the treatment choices of those who have the capacity to make them.
In the same way, respect for moral agency has given us good reason to adopt a
largely backward-looking criminal process as the primary mechanism to
deprive citizens of their liberties when they threaten harm to others.
Supporters of outpatient commitment will argue that my approach is
impervious to history. On their view, the current state of affairs in which
prisons have become the new asylums and people with mental disorders cycle
between jails, hospitals, and homeless began in the 1970s with the liberal
dismantling of mental health law. But the problem here is our broken mental
health system. Even the best studies on outpatient commitment have shown that
a court order to participate in treatment only adds value when combined with
intensive services.
Opponents of outpatient commitment will argue that these laws are a
misguided, knee-jerk response to highly publicized acts of violence by a small
number of people with mental disorders. On their view, outpatient commitment
orders are unnecessary unless the person is incompetent to refuse treatment.
However, the fact that an event is statistically rare does not mean that it is not
cause for concern. Our cognitive and volitional capacities for moral
responsibility lay the foundation upon which we base our claims to freedom
from preventive intervention. Yet when these capacities are so impaired that a
person can neither be deterred by the threat of hard treatment nor criminally
responsible for his actions, governments may employ preventive outpatient
commitment to protect others from harm.
There is a place for involuntary outpatient commitment, but these laws
require both adequate resources to ensure that effective services are available
and further amendment to protect the liberty interests of people with mental
disorders.
422. Lawrence O. Gostin & Kieran G. Gostin, A Broader Liberty: J.S. Mill, Paternalism and the Public’s Health, 123 PUB. HEALTH 214, 217 (2009).
423. Id. at 214.